Papers - THI THI ZIN
-
Advanced Predictive Analytics for Fetal Heart Rate Variability Using Digital Twin Integration. Reviewed
Tunn Cho Lwin, Thi Thi Zin, Pyke Tin, E. Kino, T. Ikenoue
Sensors (Basel, Switzerland) 25 ( 5 ) 2025.2
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal)
DOI： 10.3390/s25051469
-
Optimizing black cattle tracking in complex open ranch environments using YOLOv8 embedded multi-camera system. Reviewed International journal
Su Myat Noe, Thi Thi Zin, I. Kobayashi, Pyke Tin
Scientific reports 15 ( 1 ) 6820 2025.2
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Scientific Reports
Monitoring the daily activity levels of black cattle is a crucial aspect of their well-being. The rapid advancements in artificial intelligence have transformed computer vision applications, including object detection, segmentation, and tracking. This has led to more effective and precise monitoring techniques for livestock. In modern cattle farms, video monitoring is essential for analyzing behavior, evaluating health, and predicting estrus events in precision farming. This paper introduces the novel Customized Multi-Camera Multi-Cattle Tracking (MCMCT) system. This unique approach uses four cameras to overcome the challenges of detecting and tracking black cattle in complex open ranch environments. The MCMCT system enhances a tracking-by-detection model with the YOLO v8 segmentation model as the detection backbone network to develop a precision black cattle monitoring system. Single-camera setups in real-world datasets of our open ranches, covering 23.3 m x 20 m with 55 cattle, have limitations in capturing all necessary details. Therefore, a multi-camera solution provides better coverage and more accurate behavior detection of cattle. The effectiveness of the MCMCT system is demonstrated through experimental results, with the YOLOv8-MCMCT system achieving an average Multi-Object Tracking Accuracy (MOTA) of 95.61% across 10 cases of 4 cameras at a processing speed of 30 frames per second. This high accuracy is a testament to the performance of the proposed MCMCT system. Additionally, integrating the Segment Anything Model (SAM) with YOLOv8 enhances the system’s capability by automating cattle mask region extraction, reducing the need for manual labeling. Comparative analysis with state-of-the-art deep learning-based tracking methods, including Bot-sort, Byte-track, and OC-sort, further highlights the MCMCT’s performance in multi-cattle tracking within complex natural scenes. The advanced algorithms and capabilities of the MCMCT system make it a valuable tool for non-contact automatic livestock monitoring in precision cattle farming. Its adaptability ensures effective performance across varied ranch environments without extensive retraining. This research significantly contributes to livestock monitoring, offering a robust solution for tracking black cattle and enhancing overall agricultural efficiency and management.
-
Enhancing Fetal Monitoring through Digital Twin Technology and Entropy-Based Fetal Heart Rate Variability Analysis Reviewed International journal
Tunn Cho Lwin, Thi Thi Zin, Pyae Phyo Kyaw, Pyke Tin, E. Kino and T. Ikenoue
International Journal of Innovative Computing, Information and Control 21 ( 1 ) 185 - 196 2025.2
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：International Journal of Innovative Computing, Information and Control
In fetal healthcare, Digital Twin Technology (DTT) offers a powerful tool for simulating fetal physiological conditions, enabling continuous, real-time monitoring and predictive analysis. This study explores the integration of DTT with entropy-based analysis of fetal heart rate variability (FHRV) to enhance fetal monitoring. Utilizing a dataset of 585 fetal electrocardiogram (ECG) recordings collected via scalp electrode monitoring during delivery, we computed entropy measures such as Markov entropy and multiscale entropy to assess fetal status. The results demonstrate that these entropy measures provide significant information regarding fetal well-being status. Moreover, the calculated entropy values correlate strongly with umbilical cord blood gas parameters. This correlation suggests that entropy-based FHRV analysis, combined with DTT, can serve as an effective and reliable method for improving the accuracy of fetal health monitoring and predicting fetal well-being as delivery approaches.
-
Automated Cattle Monitoring System for Calving Time Prediction Using Trajectory Data Embedded Time Series Analysis Reviewed International journal
Wai Hnin Eaindrar Mg, Thi Thi Zin, Pyke Tin, M. Aikawa, K. Honkawa, Y. Horii
IEEE Open Journal of the Industrial Electronics Society 6 1 - 19 2025.1
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：IEEE Open Journal of the Industrial Electronics Society
This research introduces an automated system for cattle monitoring and calving time prediction, utilizing trajectory data embedded with time-series analysis. Designed for large-scale farms, our system offers continuous 12-hour monitoring, ensuring precise capture of cattle movements. By utilizing time series analysis on the trajectory data, our system predicts calving events in advance, effectively distinguishing between abnormal (requiring human assistance) and normal (not requiring assistance) for each cow. We utilized 360-degree surveillance cameras to provide comprehensive coverage without disturbing the cattle's natural behavior. We employed tailored versions of the Detectron2 and YOLOv8 models to achieve efficient and precise cattle detection, comparing their performance in terms of missed detections and false detections. For tracking, we used our Customized Tracking Algorithm (CTA), which minimizes ID switching and ensures continuous identification even in challenging conditions such as occlusions. While some ID switching errors still occur over extended tracking periods, we integrated tracking and identification to further optimize the handling of track IDs and global IDs. Our system incorporates a 4-hour forecasting of cattle movement using Euclidean Fluctuating Summation (EFS) feature combined with our custom Long Short-Term Memory (LSTM) model. Experimental results demonstrate a detection accuracy of 98.70%, tracking and identification accuracy of 99.18%, and forecasting with an average error rate of 14.07%. Furthermore, the system accurately classifies cattle as either normal or abnormal and predicts calving events a 4-hour in advance using the EFS feature, comparing its performance with various machine learning algorithms. The system's seamless integration significantly enhances farm management and animal welfare.
-
Automated system for calving time prediction and cattle classification utilizing trajectory data and movement features. Reviewed International journal
Wai Hnin Eaindrar Mg, Thi Thi Zin, Pyke Tin, M. Aikawa, K. Honkawa, Y. Horii
Scientific reports 15 ( 1 ) 2378 2025.1
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Scientific reports
Accurately predicting the calving time in cattle is essential for optimizing livestock management and ensuring animal welfare. Our research focuses on developing a robust system for calving cattle classification and calving time prediction, utilizing 12-h trajectory data for 20 cattle. Our system classifies cattle as abnormal (requiring human assistance) or normal (not requiring assistance) and predicts calving times based on their individual behaviors. We employed a tailored YOLOv8 model for efficient and precise cattle detection, effectively filtering out noise such as people and trucks. Our Customized Tracking Algorithm (CTA) maintains continuous identity tracking for each cow, enabling accurate re-identification even during occlusions. To minimize some ID switching errors over extended tracking periods, we integrated IDs optimization in the CTA utilizing Global IDs identification. We extracted and compared three total movement features for classifying cattle as abnormal or normal. For predicting calving times for each cow, we utilized and compared three cumulative movement features. Our system is fully automated, detecting and tracking all 20 cattle continuously for 12 h without manual assistance, and achieving an overall accuracy of 99%. By comparing three features derived from the trajectory tracking data for each point in a frame, we achieve 100%, 95%, 85% accuracy in classifying cattle as abnormal or normal and predict their calving times with a precision of within the next 6 h, within the next 9 h, within the next 8 h, respectively. Our system enables farmers to provide timely assistance, ensuring the health and safety of both the cow and the calf. Furthermore, it aids in optimizing resource allocation and enhancing overall farm efficiency, emphasizing the critical importance of calving time prediction in sustainable livestock farming.
-
Utilizing Behavioral Features for Predicting Calving Time Reviewed
Wai Hnin Eaindrar Mg, Pyke Tin, M. Aikawa, I. Kobayashi, Y. Horii, K. Honkawa K., Thi Thi Zin
Lecture Notes in Electrical Engineering 1321 LNEE 148 - 159 2025
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：Lecture Notes in Electrical Engineering
Accurately predicting calving time in cattle is crucial for enhancing livestock management and ensuring animal welfare. Our research presents a novel approach combining advanced computer vision and deep learning techniques to predict calving time based on cattle behavior. We employ a custom YOLOv8 model for cattle detection, achieving robust and efficient localization of individual cattle in various farm environments. Our Customized Tracking Algorithm (CTA) is utilized to maintain continuous identity tracking for each cow, allowing for precise monitoring of behavioral patterns. Feature extraction is performed using ResNet50, capturing detailed spatial characteristics of the detected cattle. These features are then organized into sequences to prepare them for temporal analysis. Finally, Custom Long Short-Term Memory (CLSTM) network is used for classification, leveraging the sequential nature of the data to predict the onset of calving with high accuracy. Our classification approach achieved an average accuracy of 94.88%. Our findings indicate promising performance from our CLSTM algorithm, accurately forecasting the remaining 3 h before calving. Through a comprehensive exploration of data collection, pre-processing, and feature engineering, our research paper establishes the foundation for training an accurate behavior model to predict calving time. Predicting calving using traditional, manual methods like observing breeding records and visual cues is complex and prone to errors, with even experts sometimes failing to make accurate predictions. Additionally, manual prediction becomes impractical and costly as farm size increases. Our automated system demonstrated a significant improvement in prediction accuracy, reducing false positives and providing timely alerts. Our proposed method demonstrates significant potential for improving the precision and reliability of calving time predictions, offering valuable insights for farm management and veterinary care.
-
Cho Nilar Phyo,Thi Thi Zin, H. Hama, Pyke Tin
Lecture Notes in Electrical Engineering 1321 LNEE 54 - 62 2025
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：Lecture Notes in Electrical Engineering
A Reinforcement Learning, a pivotal component of artificial intelligence, is employed by computers to learn intelligently. This study delves into the application of a significant mathematical concept, the Markov Decision Process (MDP), within the realm of Reinforcement Learning. The primary focus lies in the development of a proficient computer program designed to tackle healthcare issues. The proposed approach consists of two key elements: a unique decision-making framework and intelligent learning mechanisms. This process of combining two elements is leveraged to analyze patient information and ascertain optimal choices. Conceptualizing a patient’s healthcare journey as distinct states—such as physician visits (O), hospitalization (H), intensive care (I), or mortality (D)—the research formulates a Markov Chain Model. This model quantifies the transition probabilities between these states. Additionally, an auxiliary model is constructed to gauge the efficacy of decisions, encompassing factors like risk assessment and potential medication outcomes. The effectiveness of the proposed model, termed the Markov Model with Reinforcement Learning, is evaluated using real-world patient data from electronic health records. Encouragingly, the model demonstrates proficiency in predicting forthcoming healthcare events. This underscores its utility in prognosticating future developments within the healthcare domain.
-
A Study on Health Management by Behavior Analysis of Calves Reviewed
T. Nishiyama, S. Kazuhisa, M. Aikawa, I. Kobayashi, Thi Thi Zin
Lecture Notes in Electrical Engineering 1322 LNEE 144 - 151 2025
Authorship：Last author Language：English Publishing type：Research paper (international conference proceedings) Publisher：Lecture Notes in Electrical Engineering
It is important to always monitor the health of cattle, especially calves, and the frequency of observation increases with environmental changes in addition to once a day. In addition, calves tend to be more susceptible to infectious diseases because of their immature immune systems. Therefore, rearing management is extremely important. And the number of dairy cattle-keeping households and the total number of cattle are decreasing, while the number of cattle per household is increasing, indicating that management is becoming larger in scale. In this study, we proposed the development of a health management system by analyzing calf behavior using a 3D camera. Experiments were conducted at the Sumiyoshi Field of Miyazaki University to confirm the effectiveness of the proposed method.
-
Applying Digital Restoration Techniques in Preservation of Ancient Murals Using Diffusion-Based Inpainting Reviewed International coauthorship
Tint K.K.W., Mie Mie Tin, Thi Thi Zin, Pyke Tin
Lecture Notes in Electrical Engineering 1321 LNEE 398 - 407 2025
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Lecture Notes in Electrical Engineering
Nowadays, with the assistance of advanced technology, everything is constantly evolving in every sector, such as transportation, economics, medication, and so on. Also, in the sector of cultural heritage, image inpainting technology has become very popular as an effective way that allows conservators to digitally restore damaged murals without physically altering the original artwork, preserving its integrity and historical authenticity. In this paper, an image inpainting framework is proposed to restore the ancient murals of Myanmar from AD 1800–1900. The framework can be subdivided into crack removal and lacuna removal. The identification of crack and lacuna damage is automatically done with segmentation and image processing methods. Crack damage is reconstructed with pixel neighboring transfer, while lacuna reconstruction is applied with coherent transport and patch-based nearest neighbor similarity color filling methods. The accuracy is tested with the damage ratio analysis, and the experimental result demonstrates that the framework can deliver satisfactory visual results in the reconstruction process of the murals.
-
Automated Cattle Identification via Image-Based Ear Tag Recognition Reviewed
Y. Shimizu, Thi Thi Zin, M. Aikawa, I. Kobayashi
Lecture Notes in Electrical Engineering 1321 LNEE 138 - 147 2025
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：Lecture Notes in Electrical Engineering
A cattle management system using image processing technology has been proposed to reduce the labor burden of livestock farmers and improve management efficiency. To make cattle management using image processing more efficient, individual identification is necessary. Therefore, we focused on the 10-digit individual identification number given to each cow. In this study, face detection is first performed using YOLOv8, and ear regions are set from the detected face regions. The ear tag region is extracted from the ear region using color information, and the individual identification number on the ear tag is read. The identification result of the ear tag number may not be correctly identified due to the fact that a part of the ear tag is not captured due to the movement of the face or ear. Therefore, we proposed a mechanism to give a confidence score to the character identification results, update the ear tag number identification results, and obtain the ear tag number correctly, and confirmed its effectiveness through a demonstration experiment.
-
Automatic Body Temperature Detection in Calves and Alarm System Using Thermographic Camera Reviewed
Aung Si Thu Moe, Thi Thi Zin, M. Aikawa, I. Kobayashi
Lecture Notes in Electrical Engineering 1321 LNEE 190 - 198 2025
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：Lecture Notes in Electrical Engineering
The health monitoring of cows is crucial in livestock farming, particularly for calves, which are more susceptible to infectious diseases than adult cattle. This vulnerability is significantly influenced by the maturation of the calf’s immune system, rearing environment, and stress management. Traditional methods of health monitoring require substantial manpower, which can be impractical and inefficient. This paper proposes an automatic body temperature detection system for calves using thermal imaging. Leveraging the capabilities of thermal images, which are widely used in security, medical, and industrial applications, this system aims to identify and monitor the body temperature of calves. By detecting the head and eyes of the calf and extracting temperature data through the Detectron2 object detection method, the system can provide timely notifications to farmers or veterinarians. The overall average detection rate of head and eye regions was 94.5%. This approach enhances the efficiency of livestock health management, reducing the reliance on manual labor and enabling early detection of health issues.
-
Cattle Lameness Classification Using Cattle Back Depth Information Reviewed
San Chain Tun, Pyke Tin, M. Aikawa, I. Kobayashi, Thi Thi Zin
Lecture Notes in Electrical Engineering 1321 LNEE 160 - 170 2025
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：Lecture Notes in Electrical Engineering
The livestock industry plays a crucial role in sustaining agricultural production and rural economies. Monitoring cattle health, however, presents significant challenges on large farms where traditional methods require diagnosing each animal individually. Lameness is a major issue affecting cattle health, leading to decreased production performance on many farms. Timely detection of lameness is essential for providing effective early treatment. In this study, we propose a system using specialized depth cameras to monitor and analyze cattle back information for classifying lameness scores. We employ Detectron2 for cattle detection and segmentation, and the Intersection over Union (IOU) method for tracking, focusing solely on the cattle’s depth region. We extract various features from the cattle’s back depth data and utilize three different machine learning algorithms: K-Nearest Neighbor (KNN), Gradient Boosting, and Extra Trees for lameness score classification. The models KNN, Gradient Boosting, and Extra Trees showed strong training and validation results. Testing showed Extra Trees performing well with 88.2% morning and 89.0% evening accuracy. Our approach demonstrates the potential of depth camera in effectively classifying lameness scores, offering significant implications for livestock health management. This method not only improves the efficiency and accuracy of health monitoring in large-scale farming but also provides a practical solution for early detection and treatment of lameness, thereby enhancing overall farm productivity.
-
Cattle Lameness Detection Using Leg Region Keypoints from a Single RGB Camera Reviewed
Bo Bo Myint, Thi Thi Zin, M. Aikawa, I. Kobayashi, Pyke Tin
Lecture Notes in Electrical Engineering 1321 LNEE 180 - 189 2025
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：Lecture Notes in Electrical Engineering
The recent rise of machine learning and deep learning has significantly impacted the field of computer vision, particularly in tasks like object detection, object tracking, and classification. This surge in interest has underscored the critical role of feature extraction as a foundational step in these machine learning pipelines. Our research focuses on applying feature extraction techniques to a cattle lameness dataset. We specifically extract features related to the movement of key points on cattle legs across a sequence of video frames. By analyzing the variations in these points, we aim to identify features that can efficiently differentiate between lame and no lame cattle using popular machine learning algorithms. All four classifiers achieved strong testing accuracy above 75%, with SVM excelling at over 84%.
-
Cow’s Back Surface Segmentation of Point-Cloud Image Using PointNet++ for Individual Identification Reviewed
Pyae Phyo Kyaw, Pyke Tin, M. Aikawa, I. Kobayashi, Thi Thi Zin
Lecture Notes in Electrical Engineering 1321 LNEE 199 - 209 2025
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：Lecture Notes in Electrical Engineering
An automatic cow health monitoring and management system cannot function effectively without an identification system in livestock farming. While 2D image-based computer vision currently achieves high accuracy in cow identification, its effectiveness can be significantly decreased by changes in lighting, environmental factors, and other limitations. To address these limitations, an identification system based on point-cloud images will be developed by using a combination of 3D TOF camera and 2D RGB camera. This system includes detection and segmentation stage, feature extraction stage, and identification stage. In this study, I focus on detecting and segmenting of cow back surface region from a point-cloud image using the PointNet+ + algorithm. Two segmentation models are trained and compared based on single-scale grouping (SSG) and multi-scale grouping (MSG) features. The extracted cow back surface region offers a rich set of features valuable for several applications, including individual cow identification, lameness detection, and body condition scoring.
-
Evaluation of Body Condition Score for Walking Dairy Cows Using 3D Camera Reviewed
M. Chikunami, Thi Thi Zin, M. Aikawa, I. Kobayashi
Lecture Notes in Electrical Engineering 1322 LNEE 63 - 72 2025
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：Lecture Notes in Electrical Engineering
Body Condition Score (BCS) is an important index for assessing body fat accumulation in cattle and plays a crucial role in managing cattle productivity, feeding efficiency, and overall health. Currently, BCS evaluations predominantly rely on visual assessment and palpation by specialized personnel, which is time-consuming and labour-intensive. Consequently, many farms refrain from utilizing BCS for cattle management. Previous studies have focused on BCS evaluation of stationary dairy cows in rotary parlours, but this approach is not feasible for small and medium-sized livestock producers lacking such facilities. To enable BCS management for dairy cows on any farm, we propose a system utilizing image processing technology for evaluating cows while walking. In this study, we capture images using three 3D cameras and construct an estimation model using feature extraction and multiple regression analysis. This model allowed the evaluation of cows with large BCS within an error margin of 0.25. Our findings suggest that this approach can significantly streamline BCS evaluation, making it accessible and practical for a broader range of dairy farms.
-
From Vision to Vocabulary: A Multimodal Approach to Detect and Track Black Cattle Behaviors Reviewed
Su Myat Noe, Thi Thi Zin, Pyke Tin, I. Kobayashi
Lecture Notes in Electrical Engineering 1321 LNEE 171 - 179 2025
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：Lecture Notes in Electrical Engineering
This paper investigates the potential of recent image-text foundation models for classifying black cattle mounting behavior without fine-tuning. Our approach begins with the detection and tracking of each individual black cattle using deep learning-based, fine-tuned YOLOv9 and Deep OC SORT tracking. Once completed, we employ zero-shot approaches, explicitly utilizing the multi-modal Large Language and Vision Alignment (LLAVA) and Large Language Model Meta AI (LLaMA) models. These models integrate visual and linguistic information seamlessly, enabling us to leverage pre-trained knowledge to analyze and understand black cattle behavior directly from images and accompanying text descriptions. By utilizing zero-shot learning, we can bypass the resource-intensive process of model fine-tuning, making it a highly efficient approach for behavior classification. Our approach highlights the robustness and flexibility of multimodal foundation models like LLAVA and LLaMA in handling complex tasks in the agricultural domain, demonstrating their potential for broader applications without requiring extensive retraining on specific datasets. Through our experiments, we showcase the accuracy and efficiency of this zero-shot multimodal approach, providing valuable insights into black cattle mounting behavior that can enhance livestock management and monitoring practices. We introduced a novel cattle dataset tailored for this purpose. We achieved high detection accuracy with a mAP of 0.9856% using our fine-tuned YOLOv9 model and an average tracking accuracy of 94.79% across five videos. The overall accuracy of our integrated system demonstrates its efficacy in accurately classifying and tracking cattle behaviors, underscoring the potential of zero-shot learning models in precision agriculture.
-
T. Ishikawa, Thi Thi Zin, M. Aikawa, I. Kobayashi
Lecture Notes in Electrical Engineering 1322 LNEE 134 - 143 2025
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：Lecture Notes in Electrical Engineering
Rumination is a critical indicator of a cow’s physiological state, making it a valuable metric for managing cow health and predicting calving. Traditional human observation of rumination behaviour is time consuming and impractical for continuous monitoring, and sensor-based identification can be stressful for the cows due to the need for attachment. To address these challenges, this study introduces a non-contact method for identifying cow rumination. The proposed approach involves capturing video footage of multiple cows from above, recognizing cow regions, and validating the method’s effectiveness through experiments. Specifically, we utilize optical flow and frame-to-frame subtraction methods to extract moving cow regions from the recorded videos. From this data, we derive 11 features and employ a Support Vector Machine (SVM) for classification. Training the SVM with label features resulted in a test data identification accuracy of approximately 60%.
-
Research on Individual Identification of Walking Cows Using a 3D Camera Reviewed
Y. Shiihara, Thi Thi Zin, M. Aikawa, I. Kobayashi
Lecture Notes in Electrical Engineering 1322 LNEE 73 - 83 2025
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Lecture Notes in Electrical Engineering
This research focuses on identifying individual cows based on features from their back regions captured by 3D cameras, aiming to enhance management efficiency and reduce labor burdens. The methodology involves using 3D cameras to collect data on cows walking through a milking parlor. The captured data is processed to extract specific features such as pixel counts and volume from the cow’s back region. Various classification methods, including SVM, k-NN, decision trees, and random forests, are employed to identify individual cows. Experimental results demonstrated that the random forest classifier achieved the highest accuracy at 95%, outperforming other methods. The study highlights the limitations of current RFID-based systems, such as cost and stress on animals, and presents a non-contact alternative that reduces labor and improves accuracy.
-
AI-enhanced real-time cattle identification system through tracking across various environments Reviewed International journal
Su Larb Mon, T. Onizuka, Pyke Tin, M. Aikawa, I. Kobayashi, Thi Thi Zin
Scientific Reports 14 ( 1 ) 17779 2024.12
Authorship：Lead author, Last author Language：English Publishing type：Research paper (scientific journal) Publisher：Scientific Reports
Video-based monitoring is essential nowadays in cattle farm management systems for automated evaluation of cow health, encompassing body condition scores, lameness detection, calving events, and other factors. In order to efficiently monitor the well-being of each individual animal, it is vital to automatically identify them in real time. Although there are various techniques available for cattle identification, a significant number of them depend on radio frequency or visible ear tags, which are prone to being lost or damaged. This can result in financial difficulties for farmers. Therefore, this paper presents a novel method for tracking and identifying the cattle with an RGB image-based camera. As a first step, to detect the cattle in the video, we employ the YOLOv8 (You Only Look Once) model. The sample data contains the raw video that was recorded with the cameras that were installed at above from the designated lane used by cattle after the milk production process and above from the rotating milking parlor. As a second step, the detected cattle are continuously tracked and assigned unique local IDs. The tracked images of each individual cattle are then stored in individual folders according to their respective IDs, facilitating the identification process. The images of each folder will be the features which are extracted using a feature extractor called VGG (Visual Geometry Group). After feature extraction task, as a final step, the SVM (Support Vector Machine) identifier for cattle identification will be used to get the identified ID of the cattle. The final ID of a cattle is determined based on the maximum identified output ID from the tracked images of that particular animal. The outcomes of this paper will act as proof of the concept for the use of combining VGG features with SVM is an effective and promising approach for an automatic cattle identification system
-
ENHANCING FACTORY EFFICIENCY: DATA-DRIVEN ANALYSIS WITH WORKER DETECTION AND TRACKING SYSTEM Reviewed International journal
I. Hidaka, S. Inoue, T. Ishikawa, H. Tamura, Thi Thi Zin
ICIC Express Letters 18 ( 12 ) 1327 - 1337 2024.12
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters
As Japan’s population continues to decline, one challenge for small and medium-sized enterprises (SMEs) is the decline in productivity due to a shortage of employees. As a result, many small- and medium-scale factories are introducing AI and IoT to automate their operations so that they can handle them with fewer employees. However, some factories are not automating because it is more economical for employees to do the work directly. Such companies can produce more economic benefits than they do now with less labor by eliminating waste from their current work and improving work efficiency. Therefore, in this paper, we propose a system to detect and track workers in a factory using a 4K camera to obtain trajectories (lines of movement) of each work group and acquire data to be used for improving work efficiency. To determine the work group, markers are attached to the top of workers’ helmets. The proposed system is simulated with a dataset taken by us in a real factory, and the average accuracy of helmet detection and tracking is 92.9% and 86.0%, respectively. The proposed system allows us to visually see the trajectory of workers, which will lead to decision making on staffing and work process changes and improve work efficiency.
-
Predictive modeling of cattle calving time emphasizing abnormal and normal cases by using posture analysis Reviewed International journal
May Phyu Khin, Pyke Tin, Y. Horii, Thi Thi Zin
Scientific Reports 14 ( 1 ) 31871 2024.12
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Scientific Reports
Accurate calving time prediction plays a critical role in ensuring the well-being of both mother and calf during parturition. Challenges during the calving process, particularly in abnormal cases, often necessitate human intervention to prevent potentially fatal outcomes. This study proposes a novel system for automated prediction of normal and abnormal cattle calving cases based on posture analysis. By analyzing changes in posture and identifying specific posture types exhibited by cattle, the system aims to provide early warnings of impending calving events, enabling timely intervention and risk mitigation measures. Leveraging advanced computer vision techniques, particularly the Mask R-CNN from the Detectron2 detection and the YOLOv8-pose classification method known for their efficient training time and overall accuracy, the system analyzes the frequency of posture changes and key postures like sitting, standing, feeding, sitting with extended legs, and tail-raised to predict calving cases with high precision. We discovered that the “sitting with leg extended” posture is a crucial indicator for abnormal calving events. By incorporating this posture into the classification process, the system aims to achieve high accuracy in predicting both normal and abnormal calving timeframes. Additionally, the system differentiates between normal and abnormal calving patterns by analyzing posture sequences leading up to parturition, focusing on timeframes such as 30 min, 1 h, and 2 h pre-calving. This comprehensive analysis aids in identifying potential calving complications and enables the implementation of proactive management strategies. By offering insights into optimal methods for predicting specific postures and optimizing calving time management practices, this research contributes to the field of precision livestock farming, ultimately enhancing animal welfare and reducing calving-related risks.
-
Unobtrusive Elderly Action Recognition with Transitions Using CNN-RNN Reviewed
Ye Htet, Thi Thi Zin, H. Tamura, K. Kondo, E. Chosa
Journal of Signal Processing 28 ( 6 ) 315 - 319 2024.11
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：信号処理学会
This study addresses the efficient recognition of elderly people's daily actions, emphasizing transition states, using privacy-preserving depth data and deep learning algorithms. Stereo-depth cameras collect data from an elder care center, ensuring privacy by capturing only depth information without revealing identifiable details. The research investigates spatial and temporal features in movement patterns by employing a Convolutional Neural Network (CNN) for transfer learning on segmented person image sequences to extract spatial features, while a Recurrent Neural Network (RNN) decoder extracts temporal features. The proposed study evaluated various CNN and RNN integrated architectures, assessing algorithmic performance on real-world data from three elderly participants. Experimental outcomes reveal the best model achieving 95% overall accuracy for all actions and an average accuracy of over 80% for classifying transition states. Beyond accuracy, comprehensive evaluation includes precision, recall, and F1-score, offering a thorough assessment of the developed algorithm's practical effectiveness on real-world data.
DOI： 10.2299/jsp.28.315
-
A STUDY OF THE EARLY DETECTION OF OTITIS MEDIA IN CALVES WITH TWO TYPES OF CAMERAS Reviewed International journal
T. Nishiyama, K. Shiiya, M. Aikawa, I. Kobayashi, Thi Thi Zin
ICIC Express Letters, Part B: Applications 15 ( 11 ) 1183 - 1191 2024.11
Authorship：Last author Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
Calves tend to be more susceptible to infections than adult cattle. This may be due to a less mature immune system, stress from the rearing environment, and other factors. Early detection of disease can help prevent deterioration and the spread of infection. Therefore, in this study, we proposed to investigate the early detection of mycoplasma otitis media in a non-contact manner using RGB and thermal imaging cameras. We then conducted experiments at the Sumiyoshi Field of the University of Miyazaki to confirm the effectiveness of the proposed method.
-
A 3D CAMERA APPROACH TO EVALUATING BODY CONDITION SCORE IN WALKING DAIRY COWS Reviewed International journal
M. Chikunami, Thi Thi Zin, M. Aikawa, I. Kobayashi
ICIC Express Letters, Part B: Applications 15 ( 10 ) 1089 - 1097 2024.10
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
Body Condition Score (BCS) is an important index for assessing body fat accumulation in cattle and plays a crucial role in managing cattle productivity, feeding efficiency, and overall health. Currently, BCS evaluations predominantly rely on visual assessment and palpation by specialized personnel, which is time-consuming and laborintensive. Consequently, many farms refrain from utilizing BCS for cattle management. Previous studies have focused on BCS evaluation of stationary dairy cows in rotary parlors, but this approach is not feasible for small and medium-sized livestock producers lacking such facilities. To enable BCS management for dairy cows on any farm, we propose a system utilizing image processing technology for evaluating cows while walking. In this system, 3D cameras are employed to capture images, and an evaluation model is constructed using feature extraction and multiple regression analysis. This model allowed the evaluation of cows with large BCS within an error margin of 0.25.
-
Fusion of Strategic Queueing Theory and AI for Smart City Telecommunication System Reviewed International journal
Thi Thi Zin, Aung Si Thu Moe, Cho Nilar Phyo, Pyke Tin
Proceedings - 2024 IEEE 21st International Conference on Mobile Ad-Hoc and Smart Systems, MASS 2024 653 - 657 2024.10
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：Proceedings - 2024 IEEE 21st International Conference on Mobile Ad-Hoc and Smart Systems, MASS 2024
This paper explores the innovative intersection of queueing theory and artificial intelligence (AI), addressing emerging challenges for Smart City Telecommunication System. We propose a novel approach that integrates queueing theory with AI methodologies, particularly artificial neural networks (ANNs), to optimize strategic parameters in Markov Decision Process (MDP) problems in communication systems. These parameters include server numbers, customer waiting times, queue lengths, and other critical metrics. Our fusion approach demonstrates high efficacy and indicates significant advancements in applying machine learning to complex queueing theory issues. The study underscores the practical applications of this integration across various domains. We provide real-world examples illustrating the strategic use of AI-enhanced queueing models in improving user experiences and optimizing system efficiencies. Our discussions cover the benefits of intelligent resource allocation, dynamic load balancing, and adaptive service time optimization. We extend our exploration through specific case studies demonstrating the efficacy of this integration in industries such as telecommunications, healthcare, and transportation. For instance, in telecommunications, AI-driven queue management systems can dynamically allocate bandwidth to ensure optimal service quality. In healthcare, queueing models enhanced by AI can improve patient flow and reduce wait times, potentially leading to better healthcare outcomes. In this paper, we investigate the application of deep learning models to queueing theory, specifically focusing on [mention specific problem or application, e.g., 'forecasting queue lengths in real-time systems']. We implement and evaluate several deep learning architectures to determine their effectiveness in modeling and predicting queue dynamics. The paper concludes by emphasizing the necessity for continued interdisciplinary research, encouraging collaboration between AI experts and queueing theorists. This synergy is essential for unlocking new potential and addressing the increasingly complex challenges posed by modern systems. By fostering such collaboration, we anticipate significant breakthroughs that will transform both fields, leading to more efficient, adaptive, and intelligent systems capable of meeting future demands across various sectors.
-
Analyzing Parameter Patterns in YOLOv5-based Elderly Person Detection Across Variations of Data Reviewed International journal
Ye Htet, Thi Thi Zin, Pyke Tin, H. Tamura, K. Kondo, S. Watanabe, E. Chosa
Proceedings - 2024 IEEE 21st International Conference on Mobile Ad-Hoc and Smart Systems, MASS 2024 629 - 634 2024.10
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：Proceedings - 2024 IEEE 21st International Conference on Mobile Ad-Hoc and Smart Systems, MASS 2024
This study investigates the impact of data variations on parameter patterns within a YOLOv5-based elderly person detection model. We explore how changes in camera settings and environmental factors influence the model's parameters. Our research questions focus on how these variations affect parameter patterns, model robustness, and generalizability. We aim to identify the parameters and layers of the model most susceptible to variations and develop strategies to improve the model's performance across various datasets. The experiment involves collecting data from eight elderly participants in real-life elder care facility settings. During data acquisition, only depth images are recorded using stereo-depth cameras to protect privacy. After that, we train individual YOLOv5 models for each dataset through hyperparameter tuning and transfer learning. Optimal hyperparameters and the sensitive convolutional layer of each model are then compared. Class Activation Maps (CAMs) are utilized to visualize the network's focus, followed by analysis of weight distributions and correlation to identify parameter patterns. The findings will provide valuable insights for improving elderly person detection models for smart health care and their robustness to real-world variations.
-
AUTOMATED CATTLE DETECTION USING MASK R-CNN AND IOU-BASED TRACKING WITH A SINGLE SIDE-VIEW CAMERA Reviewed International journal
Bo Bo Myint, T. Onizuka, Pyke Tin, M. Aikawa, I. Kobayashi, Thi Thi Zin
International Journal of Innovative Computing, Information and Control 20 ( 5 ) 1439 - 1447 2024.10
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：International Journal of Innovative Computing, Information and Control
In precision livestock farming, the early detection of lameness in cattle is an extremely important aspect of effective breeding management. Timely identification of lameness not only facilitates prompt and cost-efficient treatment but also plays a crucial role in avoiding possible future diseases. This study emphasizes the significance of intelligent visual perception systems for lameness detection in dairy cattle, particularly in the lane between from Milking Parlor to Cattle Barn. To address the cattle lameness issue, we employ an advanced deep learning, and image processing technique, i.e., Mask R-CNN from Detectron2 to detect and identify cattle regions for feature extraction of lameness detection. On the other hand, cattle tracking using IoU is also an important part of data accumulation for lameness classification. The results of this study contribute to ongoing efforts in precision animal husbandry and demonstrate the potential of intelligent visual recognition systems for early lameness detection.
-
Precision Livestock Tracking: Advancements in Black Cattle Monitoring for Sustainable Agriculture Reviewed
Su Myat Noe, Thi Thi Zin, Pyke Tin, I. Kobayashi
Journal of Signal Processing 28 ( 4 ) 179 - 182 2024.7
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：信号処理学会
Utilizing computer vision for animal behavior monitoring provides significant benefits by minimizing direct handling and capturing diverse traits through a single sensor. However, accurately identifying animals remains a challenge. To address this, this study introduces an innovative approach to monitor black cattle in dynamic agricultural environments to ensure their health welfare. By integrating advanced techniques like DETIC for automated labeling and YOLOv8 for real-time detection, the research emphasizes improving accuracy and robustness in tracking black cattle tracking within complex open ranch environments. Moreover, the customized ByteTrack model tailored for ranch scenarios significantly enhances cattle tracking across intricate landscapes. Achieving a mean Average Precision (mAP) of 0.901 and a Multi-Object Tracking Accuracy (MOTA) of average accuracy 92.185% of four videos, this approach appears to offer a viable resolution for conducting individual cattle behavior analysis experiments through the application of computer vision.
DOI： 10.2299/jsp.28.179
-
Identifying Black Cow Actions Using Kalman Filter Velocity and Multi-Stage Classification Reviewed
Cho Cho Aye, Thi Thi Zin, M. Aikawa, I. Kobayashi
Journal of Signal Processing 28 ( 4 ) 183 - 186 2024.7
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：信号処理学会
This study proposes an advanced camera-based monitoring system for individual black cows in closed farms. By leveraging computer vision and deep learning, the system identifies five key cow actions: eating, drinking, sitting, standing, and walking. A multi-stage approach classifies actions first as static (eating, drinking, sitting, standing) or dynamic (walking) categories based on Kalman Filter velocity information. Further classification distinguishes among four static actions. A Convolutional Neural Network (CNN) refines especially for sitting and standing. On the other hand, cow head regions and specific zone locations help distinguish eating and drinking. The system achieves an overall accuracy of 80% in long data sequences, demonstrating its potential for precision livestock farming.
DOI： 10.2299/jsp.28.183
-
Development of a real-time cattle lameness detection system using a single side-view camera Reviewed International journal
Bo Bo Myint, T. Onizuka, Pyke Tin, M. Aikawa, I. Kobayashi, Thi Thi Zin
Scientific reports 14 ( 1 ) 13734 2024.6
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Scientific Reports
Recent advancements in machine learning and deep learning have revolutionized various computer vision applications, including object detection, tracking, and classification. This research investigates the application of deep learning for cattle lameness detection in dairy farming. Our study employs image processing techniques and deep learning methods for cattle detection, tracking, and lameness classification. We utilize two powerful object detection algorithms: Mask-RCNN from Detectron2 and the popular YOLOv8. Their performance is compared to identify the most effective approach for this application. Bounding boxes are drawn around detected cattle to assign unique local IDs, enabling individual tracking and isolation throughout the video sequence. Additionally, mask regions generated by the chosen detection algorithm provide valuable data for feature extraction, which is crucial for subsequent lameness classification. The extracted cattle mask region values serve as the basis for feature extraction, capturing relevant information indicative of lameness. These features, combined with the local IDs assigned during tracking, are used to compute a lameness score for each cattle. We explore the efficacy of various established machine learning algorithms, such as Support Vector Machines (SVM), AdaBoost and so on, in analyzing the extracted lameness features. Evaluation of the proposed system was conducted across three key domains: detection, tracking, and lameness classification. Notably, the detection module employing Detectron2 achieved an impressive accuracy of 98.98%. Similarly, the tracking module attained a high accuracy of 99.50%. In lameness classification, AdaBoost emerged as the most effective algorithm, yielding the highest overall average accuracy (77.9%). Other established machine learning algorithms, including Decision Trees (DT), Support Vector Machines (SVM), and Random Forests, also demonstrated promising performance (DT: 75.32%, SVM: 75.20%, Random Forest: 74.9%). The presented approach demonstrates the successful implementation for cattle lameness detection. The proposed system has the potential to revolutionize dairy farm management by enabling early lameness detection and facilitating effective monitoring of cattle health. Our findings contribute valuable insights into the application of advanced computer vision methods for livestock health management.
-
Enhancing Fetal Heart Rate Monitoring Through Digital Twin Technology Reviewed International journal
Tunn Cho Lwin, Thi Thi Zin, Pyke Tin, T. Ikenoue, E. Kino
2024 IEEE Gaming, Entertainment, and Media Conference, GEM 2024 2024.6
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：2024 IEEE Gaming, Entertainment, and Media Conference, GEM 2024
In the era of digital twin technology (DTT), which involves creating virtual replicas of physical systems, healthcare applications have seen a surge in innovation. Fetal heart rate monitoring, a rapidly advancing field within healthcare, is benefiting greatly from the effective implementation of DTTs. Digital twin, supported by Artificial Intelligence (AI), Virtual Reality (VR), and Extended Virtuality (EV), have already demonstrated significant impact in entertainment, gaming, and media sectors. This paper explores and analyzes fetal heart rate monitoring systems during labor using digital twin technologies. During labor, there is an elevated risk of developing non-communicable diseases (NCDs) such as diabetes, cardiovascular diseases, cancer, and chronic respiratory diseases. The developmental origins of health and disease hypothesis posits that environmental conditions during fetal and early postnatal development have enduring effects on growth, structure, and metabolism, ultimately influencing long-term health and well-being. Understanding and effectively monitoring fetal heart rate dynamics during labor using digital twin technologies can provide valuable insights into maternal and fetal health. The findings underscore the value of entropy analysis in FHRV assessment, presenting a pioneering predictor of fetal health via ECG data. Moreover, it reveals the current landscape of fetal heart rate monitoring, discusses the integration of digital twins, and proposes future directions for optimizing healthcare outcomes in the digital era.
-
A Stochastics Branching Process Model for Analyzing Rumor Spreading in Social Media Networks Reviewed International journal
Thi Thi Zin, Pyke Tin, H. Hama
2024 IEEE Gaming, Entertainment, and Media Conference, GEM 2024 2024.6
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：2024 IEEE Gaming, Entertainment, and Media Conference, GEM 2024
This paper introduces a stochastics branching process model as an analytical tool to study the dynamics of rumor spreading in social media networks. Employing a specialized first-order conditional probability generating function method, we derive a second-order linear regressive equation to characterize the intricate dynamics of rumor propagation. The validation of our model's accuracy is established through a comparative analysis between numerical computations and Monte Carlo simulations. Additionally, we present a derived threshold condition for the spread control factor, also known as the reproduction rate. Numerical simulation results demonstrate that the interactions and responses within social platforms contribute to the rapid onset of rumors while simultaneously diminishing the maximum density of spreaders and the overall scale of the rumors. To provide further insights, we explore analogies between rumor-spreading models and epidemic models. In conclusion, our proposed stochastics branching process model not only enhances our understanding of rumor spreading in social media networks but also offers a valuable framework for investigating the interplay between various factors influencing the dynamics of information diffusion.
-
A Markovian Game Theoretic Framework for Analysing a Queueing System with Multiple Servers Reviewed International journal
Pyke Tin, Thi Thi Zin
2024 IEEE Gaming, Entertainment, and Media Conference, GEM 2024 2024.6
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：2024 IEEE Gaming, Entertainment, and Media Conference, GEM 2024
This paper introduces a Markov game theoretic framework designed to analyze a Markovian queueing system equipped with multiple servers. Our focus lies in modelling a message transmission system, where messages traverse various transmission options, each associated with a cost and governed by a decision process. The primary objective is to investigate the impact of cooperation and communication, or their absence, among servers. The inherent uncertainty regarding the characteristics of the available transmission alternatives is mathematically captured through a Markovian game formulation. Within this framework, we quantify the inefficiency resulting from the self-interested management of individual servers and the associated loss attributed to the decision-making process. Our analysis encompasses diverse scenarios of signaling exchange among servers, providing valuable insights into the system's behavior under varying conditions.
-
Smarter Aging: Developing a Foundational Elderly Activity Monitoring System With AI and GUI Interface Reviewed International journal
Ye Htet, Thi Thi Zin, Pyke Tin, H. Tamura, K. Kondo, S. Watanabe, E. Chosa
IEEE Access 12 74499 - 74523 2024.5
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：IEEE Access
The global rise in the elderly population, which presents challenges to healthcare systems owing to labor shortages in caregiving facilities, necessitates innovative solutions for elderly care services. Smart aging technologies such as robotic companions and digital home gadgets, offer a solution to these challenges by improving the elderly's quality of life and assisting caregivers. However, limitations in data privacy, real-time processing, and reliability often hinder the effectiveness of the existing technologies. Among these, privacy concerns are a major barrier to ensuring user trust and ethical implementation. Therefore, this study proposes a more effective approach for smart aging through elderly activity monitoring that prioritizes data privacy. The proposed system utilizes stereo depth cameras to monitor the activities of the elderly. Data were collected from real-world environments with the participation of six elderly individuals from a care center and hospital. This system focuses on recognizing common daily actions of the elderly including sitting, standing, lying down, and seated in a wheelchair. Additionally, it recognizes transition states (in-between actions such as changing from sitting to standing) that are crucial for assessing balance issues. By integrating motion information with a deep-learning architecture, the system achieved a high accuracy of 99.42% in recognizing daily actions in real-time. This high accuracy was maintained even with minimal data from new environments through transfer learning, and the adaptability of this model ensured its potential for real-world applications. For intuitive interaction between the caregivers and the system, a user-friendly graphical interface (GUI) was also designed in the proposed approach.
-
Revolutionizing Cow Welfare Monitoring: A Novel Top-View Perspective with Depth Camera-Based Lameness Classification Reviewed International journal
San Chain Tun, T. Onizuka, Pyke Tin, M. Aikawa, I. Kobayashi, Thi Thi Zin
Journal of Imaging 10 ( 3 ) 2024.3
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Journal of Imaging
This study innovates livestock health management, utilizing a top-view depth camera for accurate cow lameness detection, classification, and precise segmentation through integration with a 3D depth camera and deep learning, distinguishing it from 2D systems. It underscores the importance of early lameness detection in cattle and focuses on extracting depth data from the cow’s body, with a specific emphasis on the back region’s maximum value. Precise cow detection and tracking are achieved through the Detectron2 framework and Intersection Over Union (IOU) techniques. Across a three-day testing period, with observations conducted twice daily with varying cow populations (ranging from 56 to 64 cows per day), the study consistently achieves an impressive average detection accuracy of 99.94%. Tracking accuracy remains at 99.92% over the same observation period. Subsequently, the research extracts the cow’s depth region using binary mask images derived from detection results and original depth images. Feature extraction generates a feature vector based on maximum height measurements from the cow’s backbone area. This feature vector is utilized for classification, evaluating three classifiers: Random Forest (RF), K-Nearest Neighbor (KNN), and Decision Tree (DT). The study highlights the potential of top-view depth video cameras for accurate cow lameness detection and classification, with significant implications for livestock health management.
-
DIFFUSION-BASED INPAINTING METHODS COMPARISON WITH DAMAGE AREA REDUCTION TECHNIQUES Reviewed International coauthorship International journal
Khant Khant Win Tint, Mie Mie Tin, Thi Thi Zin, Pyke Tin
ICIC Express Letters, Part B: Applications 15 ( 3 ) 303 - 309 2024.3
Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
Ancient murals beautifully reflect the social and religious characteristics of several cultural groups in a particular historical era. Unfortunately, the irreplaceable historical murals have been damaged by both natural and human-made deterioration. Image inpainting can restore the visual appeal of a mural. Image inpainting involves repairing any damaged or missing regions. In this paper, in order to address the issue of color bias, the gray scale image undergoes an inpainting process, resulting in a lack of noticeable color differences. For the mask generation, mask is generated automatically by using thresholding. That is why it prevents over-identifying damage or missing regions by user interaction. Experiments are conducted on mural images of Po-Win-Daung, Myanmar. To assess the inpainted results without the presence of a ground truth image, the paper puts forward the idea of using the damage area reduction technique for evaluation purposes. Comparisons are carried out on directional median diffusion and coherent transport methods.
-
Customized Tracking Algorithm for Robust Cattle Detection and Tracking in Occlusion Environments Reviewed International journal
Wai Hnin Eaindrar Mg, Pyke Tin, M. Aikawa, I. Kobayashi, Y. Horii, K. Honkawa, Thi Thi Zin
Sensors 24 ( 4 ) 2024.2
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Sensors
Ensuring precise calving time prediction necessitates the adoption of an automatic and precisely accurate cattle tracking system. Nowadays, cattle tracking can be challenging due to the complexity of their environment and the potential for missed or false detections. Most existing deep-learning tracking algorithms face challenges when dealing with track-ID switch cases caused by cattle occlusion. To address these concerns, the proposed research endeavors to create an automatic cattle detection and tracking system by leveraging the remarkable capabilities of Detectron2 while embedding tailored modifications to make it even more effective and efficient for a variety of applications. Additionally, the study conducts a comprehensive comparison of eight distinct deep-learning tracking algorithms, with the objective of identifying the most optimal algorithm for achieving precise and efficient individual cattle tracking. This research focuses on tackling occlusion conditions and track-ID increment cases for miss detection. Through a comparison of various tracking algorithms, we discovered that Detectron2, coupled with our customized tracking algorithm (CTA), achieves 99% in detecting and tracking individual cows for handling occlusion challenges. Our algorithm stands out by successfully overcoming the challenges of miss detection and occlusion problems, making it highly reliable even during extended periods in a crowded calving pen.
DOI： 10.3390/s24041181
-
Digital Transformation (DX) Solution for Monitoring Mycoplasma Infectious Disease in Calves: A Worldwide Health Challenge Reviewed International journal
Cho Nilar Phyo, Pyke Tin, H. Hama, Thi Thi Zin
Lecture Notes in Electrical Engineering 1114 LNEE 218 - 226 2024.1
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：Lecture Notes in Electrical Engineering
The Mycoplasma bovis (M. bovis) is a serious threat to cattle health, resulting in significant economic losses worldwide, particularly in veal calf sector. While the disease can circulate undetected, early identification of subclinical carriers is crucial. To this end, a fully automated monitoring system for Mycoplasma Infectious Disease in Calves was proposed using digital transformation technologies and AI advances. The proposed system will consist of four stages. In the first stage, an image processing technique will be developed to automatically or manually record behavioral or physiological parameters in calves while feeding at milk feeding robots. The second stage will integrate multiple data resources, such as DX records and image data, to analyze the data for detection and diagnosis of mycoplasma infection. The third stage will employ DX and AI advances to enforce the proposed monitoring system for making accurate decisions, such as whether to treat or not and what to treat calves for. In fourth stage, some experimental results will be displayed. In conclusion, the proposed automated monitoring system will provide a valuable tool for early detection of Mycoplasma Infectious Disease in calves, leading to reduce economic losses and offer timely information to address major worldwide problem.
-
AI Driven Movement Rate Variability Analysis Around the Time of Calving Events in Cattle Reviewed International journal
Wai Hnin Eaindrar Mg, Pyke Tin, M. Aikawa, I. Kobayashi, Y. Horii, K. Honkawa, Thi Thi Zin
Lecture Notes in Electrical Engineering 1114 LNEE 227 - 237 2024.1
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：Lecture Notes in Electrical Engineering
In modern cattle management, the timely detection of cattle events is crucial for ensuring both animal welfare and farm profitability. This paper introduces an innovative approach that leverages AI-driven movement rate variability analysis to predict calving events in cattle. By harnessing advanced motion tracking technologies and machine learning algorithms, this methodology offers a non-intrusive and automated means of detecting physiological and behavioral changes associated with impending calving events. Through a comprehensive exploration of data collection, pre-processing, and feature engineering, this paper establishes the foundation for training accurate AI models. These models utilize distinct movement patterns, including changes in speed, frequency, direction, and rest behavior, as predictive indicators of calving events. Real-world validation on cattle farms underscores the practical viability of the proposed approach, demonstrating its potential to revolutionize calving event detection. By transcending traditional methods, this AI-driven solution exhibits superior accuracy and efficiency, thereby contributing to enhanced animal care, optimized farm operations, and improved economic outcomes. The paper concludes by highlighting future research avenues and underscoring the transformative implications of AI-driven movement analysis for calving event prediction in the realm of agricultural technology.
-
Cow detection and tracking system utilizing multi-feature tracking algorithm Reviewed International journal
Cho Cho Mar, Thi Thi Zin, Pyke Tin, K. Honkawa, I. Kobayashi, Y. Horii
Scientific Reports 13 ( 1 ) 17423 2023.12
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Scientific Reports
In modern cattle farm management systems, video-based monitoring has become important in analyzing the high-level behavior of cattle for monitoring their health and predicting calving for providing timely assistance. Conventionally, sensors have been used for detecting and tracking their activities. As the body-attached sensors cause stress, video cameras can be used as an alternative. However, identifying and tracking individual cattle can be difficult, especially for black and brown varieties that are so similar in appearance. Therefore, we propose a new method of using video cameras for recognizing cattle and tracking their whereabouts. In our approach, we applied a combination of deep learning and image processing techniques to build a robust system. The proposed system processes images in separate stages, namely data pre-processing, cow detection, and cow tracking. Cow detection is performed using a popular instance segmentation network. In the cow tracking stage, for successively associating each cow with the corresponding one in the next frame, we employed the following three features: cow location, appearance features, as well as recent features of the cow region. In doing so, we simply exploited the distance between two gravity center locations of the cow regions. As color and texture suitably define the appearance of an object, we analyze the most appropriate color space to extract color moment features and use a Co-occurrence Matrix (CM) for textural representation. Deep features are extracted from recent cow images using a Convolutional Neural Network (CNN features) and are also jointly applied in the tracking process to boost system performance. We also proposed a robust Multiple Object Tracking (MOT) algorithm for cow tracking by employing multiple features from the cow region. The experimental results proved that our proposed system could handle the problems of MOT and produce reliable performance.
-
A Markov-Dependent stochastic approach to modeling lactation curves in dairy cows Reviewed International journal
Thi Thi Zin, Ye Htet, Tunn Cho Lwin, Pyke Tin
Smart Agricultural Technology 6 2023.12
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Smart Agricultural Technology
The modeling of lactation curves is an essential aspect of formulating farm managerial practices in dairy cows. In this study, we propose and examine a Markov-Dependent stochastic approach to modeling lactation curves in dairy cows, with the aim of developing a model that accurately fits lactation curves for a maximum number of lactations. Specifically, we develop a special type of Gamma Type Markov Chain Model that considers the first-order linear regressive property, which makes the model more realistic and reliable. We compared the proposed model with three other models - quadratic model, mixed log function, and wood model - using various goodness of fit measures such as adjusted R2, root mean square error (RMSE), and Bayesian Information Criteria (BIC). Our results showed that lactation curve modeling using the proposed model could help set management strategies at the farm level. However, it is important to optimize the modeling process regularly before implementing these strategies to enhance productivity in dairy cows. Our study contributes to the existing literature by proposing a novel approach that accounts for Markov dependence and linear regression in modeling lactation curves, which can lead to more accurate and reliable predictions. This modeling approach has practical implications for dairy farmers who seek to maximize productivity and efficiency while minimizing costs.
-
A Study on Assessment of Falling Risk in the Elderly Using a Balance Task Reviewed International journal
K. Kamahori, Thi Thi Zin
GCCE 2023 - 2023 IEEE 12th Global Conference on Consumer Electronics 513 - 514 2023.10
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：GCCE 2023 - 2023 IEEE 12th Global Conference on Consumer Electronics
Due to the risk of serious injury or death resulting from falls among the elderly, it is crucial to detect and prevent falls proactively. Performing a fall risk assessment in advance allows healthcare professionals to identify individuals at an early stage who are at risk of falling. This paper presents a method for assessing the risk of falls in the elderly using a 3D camera during a balance task. By utilizing a 3D camera, the risk of falling can be evaluated irrespective of the patient's attire or how they are dressed. Additionally, the proposed balance task is relatively simple, resulting in a relatively low burden on the participants. Features are extracted from the recorded balance task video and analyzed using a classifier. The validation results demonstrate that this method achieved an accuracy of up to 76.5% in classification.
-
Evaluating Imputation Strategies for Handling Missing Data: A Comparative Study Reviewed International journal
Tunn Cho Lwin, San Chain Tun, Pyke Tin, Thi Thi Zin
GCCE 2023 - 2023 IEEE 12th Global Conference on Consumer Electronics 508 - 509 2023.10
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：GCCE 2023 - 2023 IEEE 12th Global Conference on Consumer Electronics
Missing data is a significant challenge across various domains of data analysis, impacting the accuracy of analysis and interpretation of underlying patterns and relationships within datasets. This study specifically focuses on two different datasets: addressing the absence of depth values in cattle backbone data captured using 3-D cameras and addressing missing data in real-time recordings of RR intervals in fetal health rate variability (FHRV) obtained from the sensors used for internal monitoring of electrocardiogram (ECG) recordings taken prior to fetal delivery. To tackle these gaps, popular time series imputation techniques, including linear interpolation, spline interpolation, and autoregressive models, are employed. The performance of each model is evaluated using the root mean square error (RMSE). This study ultimately selects the optimal model for handling the missing data which is important for data analysis research work.
-
A NON-INVASIVE METHOD FOR LAMENESS DETECTION IN DAIRY COWS USING RGB CAMERAS Reviewed International journal
T. Onizuka、Thi Thi Zin, I. Kobayashi
ICIC Express Letters, Part B: Applications 14 ( 10 ) 1107 - 1114 2023.10
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
Lameness is a major health issue affecting dairy cows, causing pain, discomfort, and abnormal movements that can lead to decreased productivity and other diseases. Early detection and treatment are crucial to prevent the development of more serious conditions. In this paper, we present a method for detecting lameness in dairy cows using an RGB camera and analyzing their walking behavior. Our proposed technique achieves an accuracy of 84.6% in classifying cows as healthy or lame. We conducted a series of real-life experiments to validate our classification results, comparing them with expert diagnoses. Our method has the potential for use in routine farming conditions to detect lameness early and improve cow welfare and productivity.
-
A Markov Chain Model for Determining the Optimal Time to Move Pregnant Cows to Individual Calving Pens Reviewed International journal
Cho Nilar Phyo, Pyke Tin, Thi Thi Zin
Sensors 23 ( 19 ) 2023.10
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Sensors
The use of individual calving pens in modern farming is widely recognized as a good practice for promoting good animal welfare during parturition. However, determining the optimal time to move a pregnant cow to a calving pen can be a management challenge. Moving cows too early may result in prolonged occupancy of the pen, while moving them too late may increase the risk of calving complications and production-related diseases. In this paper, a simple random walk type Markov Chain Model to predict the optimal time for moving periparturient cows to individual calving pens was proposed. Behavior changes such as lying time, standing time, and rumination time were analyzed using a video monitoring system, and we formulated these changes as the states of a Markov Chain with an absorbing barrier. The model showed that the first time entering an absorbing state was the optimal time for a pregnant cow to be moved to a calving pen. The proposed method was validated through a series of experiments in a real-life dairy farm, showing promising results with high accuracy.
DOI： 10.3390/s23198141
-
Efficient Segment-Anything Model for Automatic Mask Region Extraction in Livestock Monitoring Reviewed International journal
Su Myat Noe, Thi Thi Zin, Pyke Tin, I. Kobyashi
IEEE International Conference on Consumer Electronics - Berlin, ICCE-Berlin 167 - 171 2023.9
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：IEEE International Conference on Consumer Electronics - Berlin, ICCE-Berlin
This paper presents an efficient segment-anything model for automatic mask region extraction in livestock tracking. The research focuses on developing and evaluating automatic mask segmentation models for tracking black cattle. The primary contribution is a tailored extraction segmentation model for automatically extracting cattle mask regions utilizing in the livestock tracking. The methodology utilizes Segment Anything Model (SAM), Grounded SAM, Grounding Dino, YOLOv8, and DeepOCSort algorithms for detection and tracking. Experimental results demonstrate the effectiveness of the proposed approach in extracting black cattle mask regions and improving livestock tracking. Integration of YOLOv8 and DeepOCSort ensures accurate association and tracking of mask regions across frames. The findings advance livestock tracking, with applications in precision agriculture. The proposed segment-anything model serves as a valuable tool for automatic mask region extraction in foreground-background separation.
-
Markov Chain Modelling for Heart Rate Variability Analysis: Bridging Artificial Intelligence and Physiological Data Reviewed International journal
Thi Thi Zin, Pyke Tin
IEEE International Conference on Consumer Electronics - Berlin, ICCE-Berlin 163 - 166 2023.9
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：IEEE International Conference on Consumer Electronics - Berlin, ICCE-Berlin
Heart rate variability (HRV) is a vital measure that captures the variations in beat-to-beat intervals. It is a significant indicator of autonomic nervous system activity and has been linked to various health conditions. Analyzing HRV involves processing complex physiological data with inherent uncertainties. Artificial Intelligence (AI) refers to the development of intelligent machines capable of mimicking human intelligence and learning. Markov Chain theory provides a powerful framework for studying and mathematically modelling situations with random variables. In this paper, we explore the application of Markov Chain concepts to HRV analysis, treating it as an AI problem. We represent beat-to-beat intervals as a sequence of random variables, forming Markov Chain states where the current state depends on the immediate previous state. We derive the state-to-state probability transition matrix and compute the stationary distribution probabilities. Using these probabilities, we calculate the Shannon entropy measure to gain insights into heart rate variability. Furthermore, we present the utilization of real-life experimental data to illustrate the effectiveness of the proposed method. The experimental results demonstrate the promising potential of the integration of Markov Chain and AI in analyzing HRV by achieving the highest accuracy of 95%. This research a novel perspective on understanding the underlying dynamics of HRV and its implications for human health.
-
Temporal-Dependent Features Based Inter-Action Transition State Recognition for Eldercare System Reviewed International journal
Ye Htet, Thi Thi Zin, H. Tamura, K. Kondo, E. Chosa
IEEE International Conference on Consumer Electronics - Berlin, ICCE-Berlin 106 - 111 2023.9
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：IEEE International Conference on Consumer Electronics - Berlin, ICCE-Berlin
Elderly individuals are particularly vulnerable to accidents, with a significant number of incidents occurring during transition states between primitive actions such as sitting to standing and sitting to lying. This paper introduces a novel machine-learning technique in artificial intelligence, based on temporal-dependent features to assist the elderly. To ensure privacy, we employed stereo depth cameras for data acquisition from the elder care center and exclusively processed depth images. The first step of our approach involves localizing individuals using the YOLOv5 object detector. Subsequently, we employed the Segment Anything Model to segment only the person masks, excluding other areas from consideration. Temporal-dependent features were then extracted for every five frames from the subsequent person masks that enable the recognition of transition states from primitive actions. We tested various classification approaches and compared the results by defining norms and metrics. Our experimental findings demonstrated that the overall accuracy rates for classifying 2 classes and 5 classes on small segments are 91.18% and 91.67% respectively. To validate the effectiveness of our proposed method, we conducted experiments using real-life environments inside three rooms and obtained average accuracy rates of 90.17%, 97.16%, and 77.44% respectively. Overall, this model has the potential to enhance the safety and well-being of the elderly population.
-
A COMPARATIVE STUDY ON COW RECOGNITION: ANALYZING COLOUR SPACES, DISTANCE MEASURES AND DEEP NEURAL NETWORKS Reviewed International journal
Cho Cho Mar, Thi Thi Zin, Pyke Tin, K. Honkawa, I. Kobayashi, Y. Horii
ICIC Express Letters, Part B: Applications 14 ( 9 ) 993 - 1000 2023.9
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
The implementation of autonomous cattle monitoring systems is becoming increasingly important in the livestock production and dairy farming industries. These systems require robust object detection and tracking systems to enhance their performance and reliability. The recognition task plays a crucial role in building a powerful tracking system. The aim of this study is to perform the recognition task for cow recognition by analyzing and comparing different colour spaces and distance measures to select the optimal ones. We extracted colour moment features and Co-occurrence Matrix (CM) features from various colour spaces: RGB, YCbCr, XYZ, HSV, CIELab, and grey level. We compared these features using different distance measures based on their accuracy values. We also conducted experiments on different pre-trained deep neural networks to extract Convolutional Neural Network (CNN) features and compared the accuracy values of the classification results with the Support Vector Machine (SVM) method. These experiments were conducted on the cow dataset created from a continuous 2-hour video. The results demonstrate that the combination of CM features extracted from the HSV colour space, and the Manhattan distance measure produced the highest accuracy values. Furthermore, we found that using the InceptionV3 pre-trained deep neural network produced the best accuracy results when combined with the SVM classifier. These findings provide insights into optimizing cow recognition for autonomous monitoring systems.
-
PREDICTING DAIRY COW CALVING TIME USING MARKOV MONTE CARLO SIMULATION AND NAÏVE BAYES CLASSIFIER Reviewed
Thi Thi Zin, Swe Zar Maw, Pyke Tin, Y. Horii, H. Hama
ICIC Express Letters, Part B: Applications 14 ( 8 ) 877 - 888 2023.8
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
This study proposes an approach that utilizes both Markov Monte Carlo simulation and the Bayesian method to predict the time of calving events in dairy cows. Continuous video surveillance was conducted on 30 individual dairy cows 24 hours before calving. Behaviors such as lying, transition from lying to standing, standing, and transition from standing to lying were annotated for each cow between 72 to 168 hours prior to calving. The probabilities for each behavior were derived and used in Markov Monte Carlo simulations to generate behavior patterns of each cow before calving. Three types of datasets, actual, simulated, and a mixture of the two, were investigated using Naïve Bayes Classifier for prediction. The experimental results showed that the hybrid approach accurately classified the calving event, cent by cent. This approach can assist farmers and veterinarians in making informed decisions and taking appropriate actions before the calving event, ultimately improving the health and welfare of dairy cows.
-
Quantifying Body Movement Variability for Cattle Lameness Detection Using Imaging Reviewed International journal
Thi Thi Zin and Pyke Tin
ICIC Express Letters, Part B: Applications 14 ( 7 ) 735 - 74 2023.7
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
This paper proposes a framework for detecting cattle lameness by quantifying the variability of body movement using depth imaging data collected while cows walk from the milking center to the resting area. The framework identifies critical factors that determine lameness scores based on the root mean square successive differences, various types of information entropies, and geometric measures of the collected depth data. To analyze lameness status, we developed an operational simulation model that combines Monte Carlo simulation with popular probability distribution functions such as uniform, normal, Poisson, and Gamma distributions. The simulation results suggest that detection performance and the characteristics of lame and non-lame cows significantly affect body movement variability. By using real-life data, we aim to validate this conjecture in future work.
-
Artificial Intelligence Fusion in Digital Transformation Techniques for Lameness Detection in Dairy Cattle Reviewed
Thi Thi Zin, Ye Htet, San Chain Tun and Pyke Tin
International Journal of Biomedical Soft Computing and Human Sciences (IJBSCHS) 28 ( 1 ) 1 - 8 2023.7
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (scientific journal)
-
Facial Region Analysis for Individual Identification of Cows and Feeding Time Estimation Reviewed International journal
Yusei Kawagoe, Ikuo Kobayashi, Thi Thi Zin
Agriculture (Switzerland) 13 ( 5 ) 2023.5
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (scientific journal)
With the increasing number of cows per farmer in Japan, an automatic cow monitoring system is being introduced. One important aspect of such a system is the ability to identify individual cows and estimate their feeding time. In this study, we propose a method for achieving this goal through facial region analysis. We used a YOLO detector to extract the cow head region from video images captured during feeding with the head region cropped as a face region image. The face region image was used for cow identification and transfer learning was employed for identification. In the context of cow identification, transfer learning can be used to train a pre-existing deep neural network to recognize individual cows based on their unique physical characteristics, such as their head shape, markings, or ear tags. To estimate the time of feeding, we divided the feeding area into vertical strips for each cow and established a horizontal line just above the feeding materials to determine whether a cow was feeding or not by using Hough transform techniques. We tested our method using real-life data from a large farm, and the experimental results showed promise in achieving our objectives. This approach has the potential to diagnose diseases and movement disorders in cows and could provide valuable insights for farmers.
-
Validation and Discussion of Severity Evaluation and Disease Classification Using Tremor Video Reviewed International journal
T. Hayashida, T. Sugiyama, K. Sakai, N. Ishii, H. Mochizuki, Thi Thi Zin
Electronics (Switzerland) 12 ( 7 ) 2023.4
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Electronics (Switzerland)
A tremor is a significant symptom of Parkinson’s disease, but it can also be a characteristic of essential tremor, thereby hampering even specialists’ ability to differentiate between the two. This study proposes a system that leverages a single RGB camera to evaluate tremor severity and support the differential diagnosis of Parkinson’s disease and essential tremor. The system captures motor symptoms, performs time–frequency analysis using wavelet transforms, and classifies severity and disease using linear classification models. The results showed an accuracy rate of 0.56 for disease classification and 0.50 for severity classification (with an acceptable accuracy rate of 0.96). The analysis indicated that there was a low level of correlation between disease and each feature and a moderate correlation (about 0.6) between severity and each feature. Based on these results, this study recommends classifying severity with a linear model and disease with a nonlinear model to obtain improved accuracy.
-
Comparing State-of-the-Art Deep Learning Algorithms for the Automated Detection and Tracking of Black Cattle Reviewed International journal
Su Myat Noe, Thi Thi Zin, Pyke Tin, I. Kobayashi
Sensors (Basel) 23 ( 1 ) 2023.1
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal)
-
Cattle Face Detection with Ear Tags Using YOLOv5 Model Reviewed International journal
Wai Hnin Eaindrar Mg, Thi Thi Zin
ICIC Express Letters, Part B: Applications 14 ( 1 ) 65 - 72 2023.1
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
With the increasing population’s need for meat and requirements for high food quality, the livestock industry is developing from small-scale and subsistence farming towards intensive and specialized grazing. Cattle monitoring and management system is crucial to be registered for breeding association, food quality tracing, disease prevention and control and fake insurance claims. This research presents cattle face detection with their ear tags’ names by applying light-weight YOLOv5 (You Only Look Once) model. This research is intent to the farmers who can not only monitor and manage the cattle conditions at the farm. The proposed system was trained to get the best accuracy model. The accuracy of the proposed model achieves up to 99.4% for four surveillance cameras.
-
Predicting Calving Time of Dairy Cows by Autoregressive Integrated Moving Average (ARIMA) Model and Exponential Smoothing Model Reviewed International journal
Tunn Cho Lwin, Thi Thi Zin and Pyke Tin
ICIC Express Letters, Part B: Applications 14 ( 1 ) 73 - 79 2023.1
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
Calving time prediction is an important factor in dairy farming. The careful monitoring of cows can help to decrease the loss of calf rates during the calving time; moreover, to know the exact time of birth is crucial to make sure timely assistance. However, direct visual observation is time-wasting for observers, and the continuous presence of observers during calving time may disturb cows. In this study, the recording from video cameras and counting the number of standing-to-lying and lying-to-standing transitions of 25 cows from a large farm at Oita prefecture, Japan, before 72 hours of calving time are applied. To be specific, we model the number of changes in behaviors of standing and lying as a time series in hourly basis. The time series approaches, namely the exponential distribution probability, autoregressive integrated moving average (ARIMA) model, and double exponential smoothing (DES) model, are applied to predicting the calving time and the root mean square error (RMSE) is used to check the accuracy and error value of the experiment. By investigating the changes in behavior patterns a few days before the calving events, the proposed method can predict accurately the time of occurrence of calving events by the developed ARIMA (2,0,1) model. Therefore, the developed model can be used to estimate the calving time which has significantly positive impact for livestock specialists.
-
Comparing State-of-the-Art Deep Learning Algorithms for the Automated Detection and Tracking of Black Cattle Reviewed International journal
Su Myat Noe, Thi Thi Zin, Pyke Tin, I. Kobayashi
Sensors 23 ( 1 ) 532 2023.1
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Sensors
Effective livestock management is critical for cattle farms in today’s competitive era of smart modern farming. To ensure farm management solutions are efficient, affordable, and scalable, the manual identification and detection of cattle are not feasible in today’s farming systems. Fortunately, automatic tracking and identification systems have greatly improved in recent years. Moreover, correctly identifying individual cows is an integral part of predicting behavior during estrus. By doing so, we can monitor a cow’s behavior, and pinpoint the right time for artificial insemination. However, most previous techniques have relied on direct observation, increasing the human workload. To overcome this problem, this paper proposes the use of state-of-the-art deep learning-based Multi-Object Tracking (MOT) algorithms for a complete system that can automatically and continuously detect and track cattle using an RGB camera. This study compares state-of-the-art MOTs, such as Deep-SORT, Strong-SORT, and customized light-weight tracking algorithms. To improve the tracking accuracy of these deep learning methods, this paper presents an enhanced re-identification approach for a black cattle dataset in Strong-SORT. For evaluating MOT by detection, the system used the YOLO v5 and v7, as a comparison with the instance segmentation model Detectron-2, to detect and classify the cattle. The high cattle-tracking accuracy with a Multi-Object Tracking Accuracy (MOTA) was 96.88%. Using these methods, the findings demonstrate a highly accurate and robust cattle tracking system, which can be applied to innovative monitoring systems for agricultural applications. The effectiveness and efficiency of the proposed system were demonstrated by analyzing a sample of video footage. The proposed method was developed to balance the trade-off between costs and management, thereby improving the productivity and profitability of dairy farms; however, this method can be adapted to other domestic species.
DOI： 10.3390/s23010532
-
Black Cow Tracking by Using Deep Learning-Based Algorithms Reviewed International journal
Cho Cho Aye, Thi Thi Zin, I. Kobayashi
ICIC Express Letters, Part B: Applications 13 ( 12 ) 1313 - 1319 2022.12
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
Raising livestock is essential in the farming industry to meet the consumer’s requirement. Livestock monitoring system is useful to monitor their health without need-ing too much manpower. Thus, livestock tracking becomes one of the vital parts of livestock monitoring system. The objective of this proposed system is to track black cows based on detected features. Here, the YOLOv5 (You Only Look Once) model was used in detection phase to detect cow regions and Deep SORT (Simple Online Real-time Tracking) was applied to tracking the target cows in every consecutive frame. In Deep SORT, it includes appearance feature model to recognize cow’s visual appearance such as shape, size, and pose. The proposed system was best trained by adopting transfer learning method. The detection model achieves an accuracy of 0.995 mAP@0.5 whereas the tracking model gets the performance results in video-1 and video-2 with 99.4% and 98.9%, respectively.
-
Cattle Pose Classification System Using DeepLabCut and SVM Model Reviewed International journal
May Phyu Khin, Thi Thi Zin, Cho Cho Mar, Pyke Tin, Y. Horii
GCCE 2022 - 2022 IEEE 11th Global Conference on Consumer Electronics 494 - 495 2022.10
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：GCCE 2022 - 2022 IEEE 11th Global Conference on Consumer Electronics
This paper proposes the cattle pose classification system by using video-based tracking result. The proposed system is composed of two processes namely feature extraction and classification. In the feature extraction, we employ DeepLabCut network to obtain location feature points which are to be combined with cattle bounding box region values. For classification process, the SVM (Support Vector Machine) classifier will be used. To confirm the proposed method, we tested some experimental results by using the video sequences taken in some real-life dairy farms and classify six poses such as 'standing', 'sitting', 'eating', 'drinking', 'sitting with leg extend' and 'tail raised'. We got average 88.75% accuracy for all poses.
-
Video-based Automatic Cattle Identification System Reviewed International journal
Su Larb Mon, Thi Thi Zin, Pyke Tin, I. Kobayashi
GCCE 2022 - 2022 IEEE 11th Global Conference on Consumer Electronics 490 - 491 2022.10
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：GCCE 2022 - 2022 IEEE 11th Global Conference on Consumer Electronics
In this paper, we propose a method to identify the cattle by using video sequences. In order to do so, we first collect 360-degree top-view video sequences to form dataset. The proposed system is composed of two parts: cattle detection and cattle identification. In the detection process, we utilize YOLOv5(Y ou Only Look Once) model to detect the cattle region in the lane. In this stage, cattle's location and region information are extracted and the cropped images of detected cattle regions are saved for the next stage. We then apply Convolutional Neural Network model (VGG16) to extract the features which will be used to identify individual cattle. For the classification, the proposed system used two supervised machine learning methods, Random Forest and SVM (Support Vector Machine). The accuracy of Random Forest is 98.5% and the accuracy of SVM is 99.6%. After comparing the accuracy rate of two methods, SVM get the better accuracy result. The proposed system achieved the accuracy of over 90% for both cattle detection and identification.
-
Cow Lameness Detection Using Depth Image Analysis Reviewed International journal
San Chain Tun, Thi Thi Zin, Pyke Tin, I. Kobayashi
GCCE 2022 - 2022 IEEE 11th Global Conference on Consumer Electronics 492 - 493 2022.10
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：GCCE 2022 - 2022 IEEE 11th Global Conference on Consumer Electronics
In this paper, we introduce to detect cow lameness by using a depth video camera from the top view position. To classify cow lameness, we first extract the sequences of depth value of the cow body region and the maximum value of the back cow area. Then, we will find the average from the maximum height values of the cow backbone area. By using the average values as a feature vector, we classify the cow lameness with the aid of the Support Vector Machine (SVM). To confirm, we perform some experiments by using depth camera images on a real-life dairy farm. The experimental result shows that our proposed method is promising.
-
Color Space Conversion Technique for Cattle Region Extraction with Application to Estrus Detection Reviewed International journal
Y. Hashimoto, H. Hama, Thi Thi Zin
ICIC Express Letters 16 ( 10 ) 1095 - 1100 2022.10
Authorship：Last author Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters
In modern dairy and beef farming with no exception in Japanese livestock industry, an accurate and timely estrus (heat) detection is an important and key factor in efficient and profitable reproductive management performance of cattle herd. Failure in heat detection is costly to the producer and it is considered the critical component of reproductive management. Among many estrus behaviors, visual postures of an individual cow can be successfully recognized and utilized for heat detection. In this aspect, to achieve precise identification and to obtain individual cattle information, extracting cattle region from its background is the fundamental and important step. In general, the inter-frame difference and the background subtraction are widely known as methods to detect moving objects in video images. However, these conventional methods do not work well in Japanese black cattle environments, due to their slow movements. At the same time, since the skin is similar to soil in color, region extraction is not so easy, even if background subtraction is used. Therefore, in this paper, we propose a new method for extracting cattle regions using color space conversion. The proposed method is able to automatically extract cattle regions and tracked cattle from change of the gravity center of the extracted cattle regions. Experimental results show that our approach is effective and promising with high accuracy.
-
HMM-Based Action Recognition System for Elderly Healthcare by Colorizing Depth Map Reviewed International journal
Ye Htet, Thi Thi Zin, Pyke Tin, H. Tamura, K. Kondo, E. Chosa
International Journal of Environmental Research and Public Health 19 ( 19 ) 2022.10
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：International Journal of Environmental Research and Public Health
Addressing the problems facing the elderly, whether living independently or in managed care facilities, is considered one of the most important applications for action recognition research. However, existing systems are not ready for automation, or for effective use in continuous operation. Therefore, we have developed theoretical and practical foundations for a new real-time action recognition system. This system is based on Hidden Markov Model (HMM) along with colorizing depth maps. The use of depth cameras provides privacy protection. Colorizing depth images in the hue color space enables compressing and visualizing depth data, and detecting persons. The specific detector used for person detection is You Look Only Once (YOLOv5). Appearance and motion features are extracted from depth map sequences and are represented with a Histogram of Oriented Gradients (HOG). These HOG feature vectors are transformed as the observation sequences and then fed into the HMM. Finally, the Viterbi Algorithm is applied to recognize the sequential actions. This system has been tested on real-world data featuring three participants in a care center. We tried out three combinations of HMM with classification algorithms and found that a fusion with Support Vector Machine (SVM) had the best average results, achieving an accuracy rate (84.04%).
-
A Special Type of Markov Branching Process Model for the Novel Coronavirus (Covid-19) Outbreak Reviewed International journal
Thi Thi Zin, Pyke Tin, H. Hama
International Journal of Innovative Computing, Information and Control 18 ( 4 ) 1339 - 1346 2022.8
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：International Journal of Innovative Computing, Information and Control
Mathematical modeling has been an important tool to estimate key factors of the transmission and investigate the dynamical system of evolutionary nature in epidemics. More precisely, the outbreaks of the virus or epidemiology is generally considered as an application of branching process. Therefore, in this paper, we propose a special type of Markov branching process model to examine and explore some problems of the novel Coronavirus (COVID-19) infectious disease with the aims of reducing the effective reproduction number of an infection below unity. Since the COVID-19 has been recognized as a global pandemic, we have assessed a big amount of data such as hourly contagious, hospitalized patients, recovered and deaths. However, these data are necessary to be further processed to produce useful information for people and authorities when they make an efficient and optimal decisions. In such a decision-making process, we establish a special type of Gama Markov branching process model which has been successfully applied in other research areas such as queueing and waiting lines problems, stochastic reservoir problems, inventory controls and operation research. Specifically, we develop a three parameter Gama Markov branching process model that is structured in two parts, initial and latter transmission stages, so as to provide a comprehensive view of the virus spread through basic and effective reproduction numbers respectively, along with the probability of an outbreak sizes and duration. As an illustration, we have performed some simulations based on the daily data appearing on WHO dashboard in order to analyze the first semiannual spread of the ongoing Coronavirus pandemic in the region of Myanmar. The results show that the proposed model can be utilized for the real-life applications.
-
An Intelligent Vision-Based Approach for Work Group Identification through Helmet Detection Reviewed International journal
S. Inoue, I. Hidaka, Thi Thi Zin
ICIC Express Letters, Part B: Applications 13 ( 5 ) 511 - 517 2022.5
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
Helmets are essential equipment to protect workers from danger during inspection and operation in almost all industries. There is a growing necessity of developing innovative methods to automatically monitor safety and work group identification at industry work sites. With the rapid development of artificial intelligence (AI) based image recognition technologies, computer vision-based inspections have been one of the most important industrial application areas for automation. Thus, in this paper, we propose an intelligent computer vision approach for work group identification through helmet detection by analyzing images collected from 4K camera installed overhead at work site. For this purpose, we attach a marker on the top of the worker’s helmet to detect the helmet and identify the work group. This approach is tested on our data set through simulated experiments and the average accuracy of helmet detection is 92.9%.
-
A Deep Learning Method of Edge-Based Cow Region Detection and Multiple Linear Classification Reviewed International journal
Thi Thi Zin, Saw Zay Maung Maung, Pyke Tin
ICIC Express Letters, Part B: Applications 13 ( 4 ) 405 - 412 2022.4
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
In this paper we propose a deep learning method of cow region detection and multiple linear model for classifying behaviors of pregnant cows prior to the occurrence of calving events. Dairy farm management experts and farmers have been well recognized that video monitoring to an individual cow plays an important and significant role in production and re-production processes. To be specific we can learn through video monitoring cow’s health conditions, body conditions even the occurrences of calving difficulties in times. Moreover, due to the advances in the latest computer vision and image processing algorithms, it is possible to develop a camera system that automatically detects the cow’s conditions at a low cost. The fundamental and foremost important step in the proposed system is to detect and segment the cow regions in the video sequences. After the detection process we performed a multiple linear model to classify some behaviors of the detected cows. In particular we consider four states of cow behaviors such as lying state, transition on state from lying to standing, standing state and transition state of standing to lying which are important in studying dairy cow management systems. In order to confirm the validity of our proposed method some experiments are carried out by establishing the video monitoring cameras at the maternity pens of a large dairy farm in Japan. The experimental results show that the proposed method gives an impression of promising with high accuracy.
-
Individual Identification of Cow Using Image Processing Techniques Reviewed
Y. Kawagoe, Thi Thi Zin, I. Kobayashi
2022 IEEE 4th Global Conference on Life Sciences and Technologies (LifeTech) 570 - 571 2022.3
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：LifeTech 2022 - 2022 IEEE 4th Global Conference on Life Sciences and Technologies
Cow identification has become important in recent years due to outbreaks of diseases such as bovine spongiform encephalopathy. Conventional identification methods available are not efficient, affordable, non-invasive, and cost- effective. Among them, some methods are based on biological markers, such as muzzle point matching and facial recognition. Facial images are the most common biometric characteristics used by humans to identify individuals, and they have received much attention. In this study, we used RGB camera to identify individual cow by their faces and confirmed the effectiveness of this method.
-
Action Recognition System for Senior Citizens Using Depth Image Colorization Reviewed
Ye Htet, Thi Thi Zin, H. Tamura, K. Kondo, E. Chosa
2022 IEEE 4th Global Conference on Life Sciences and Technologies (LifeTech) 494 - 495 2022.3
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：LifeTech 2022 - 2022 IEEE 4th Global Conference on Life Sciences and Technologies
This paper describes about the system which can be used at the care center for the purpose of elderly action recognition using depth camera. The depth image colorization is used for compression, visualization, and person detection process. The YOLOv5 (You Only Look Once) algorithm is used as object detector. The space-time features are extracted from depth sequences and they are recognized by linear SVM (Support Vector Machine) classifier. The random image sequences are generated for testing to recognize six actions. The results show that this system can detect the various actions with the average of 92% accuracy for different durations.
-
An Intelligent Method for Detecting Lameness in Modern Dairy Industry Reviewed
Thi Thi Zin, Moe Zet Pwint, Su Myat Noe, I. Kobayashi
2022 IEEE 4th Global Conference on Life Sciences and Technologies (LifeTech) 564 - 565 2022.3
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：LifeTech 2022 - 2022 IEEE 4th Global Conference on Life Sciences and Technologies
Lameness is one of the major welfare concerns in the modern dairy industry. In addition, lameness makes severe health and economic problems causing losses in milk production. Although there has a sizable amount of methods, it remains some worthwhile open problems. Therefore, in this paper, we propose an intelligent method for detecting the lameness of dairy cow by establishing a visual monitoring system on the laneways after milking process. We employ a technique of Mask-RCNN for cow region extraction and utilize features based on head bob patterns. Our real-life experimental results show that the proposed method has detection accuracy of 95.5% on cow's region extraction and can classify 80% of the lameness levels correctly.
-
A Deep Learning-based solution to Cattle Region Extraction for Lameness Detection Reviewed
Su Myat Noe, Thi Thi Zin, Pyke Tin, I. Kobayashi
2022 IEEE 4th Global Conference on Life Sciences and Technologies (LifeTech) 572 - 573 2022.3
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：LifeTech 2022 - 2022 IEEE 4th Global Conference on Life Sciences and Technologies
In precision livestock farming, lameness detection in cattle is particularly important for breeding management. The accurate detection of lameness is crucial for delivering effective and economical treatment and for preventing future diseases. The noticeable sign of lameness is that their speed of walking, arching their backs and drop their heads during walking. Here, we emphasis on lameness of dairy cattle by implementing the intelligent visual perception system on the laneways after milking process. Employing a deep learning technique of Mask-RCNN for cattle region detection and identification. The novelty of this work noticeably implies that deep learning instance segmentation could be effectively employed as a cattle region extraction from complex background prior to using identification and tracking.
-
A Study on Automatic Individual Identification of Wild Horses Reviewed
K. Shiiya, R. Yamada, Thi Thi Zin, I. Kobayashi
2022 IEEE 4th Global Conference on Life Sciences and Technologies (LifeTech) 492 - 493 2022.3
Language：English Publishing type：Research paper (international conference proceedings) Publisher：LifeTech 2022 - 2022 IEEE 4th Global Conference on Life Sciences and Technologies
Wild horses called "Misaki-uma"are inhabited Cape Toi in the southern part of Miyazaki Prefecture in Japan. Although wild horse, it needs health care. It is a difficult task for aging management association members to monitor the vast habitat range of horses and to identify individual during routine management. In addition, the current methods of individual identification because of contact with horses or to requiring specialized knowledge, ordinary people cannot perform it. In this paper we propose a method for automatic individual identification of wild horses without contact using an RGB camera.
-
A Study on Worker Tracking Using Camera to Improve Work Efficiency in Factories Reviewed
I. Hidaka, S. Inoue, Thi Thi Zin
2022 IEEE 4th Global Conference on Life Sciences and Technologies (LifeTech) 568 - 569 2022.3
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：LifeTech 2022 - 2022 IEEE 4th Global Conference on Life Sciences and Technologies
The population of Japan is declining every year. As the population declines, the number of employees in enterprises also decreases, and one of the issues for small and medium-sized enterprises is the decline in productivity due to a shortage of employees. It is necessary to improve work efficiency to compensate for the shortage of employees and the resulting decrease in productivity. If changes in the work process and unnecessary movements in the work process can be eliminated, it will lead to shortening of work time and improvement of work efficiency. In this paper, to improve work efficiency for factories, we aim to track the workers in the factory using a camera and show the trajectory of works.
-
Cho Cho Mar, Thi Thi Zin, I. Kobayashi, Y. Horii
2022 IEEE 4th Global Conference on Life Sciences and Technologies (LifeTech) 566 - 567 2022.3
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：LifeTech 2022 - 2022 IEEE 4th Global Conference on Life Sciences and Technologies
Cow detection and tracking system plays an important role in cattle farming and diary community to reduce expenses and workload. This research presents how the conventional image processing techniques can be combined with deep learning concepts to establish cow detection and tracking system. Specifically, we first employ a Hybrid Task Cascade (HTC) instance segmentation network for cow detection. We then built the multiple objects tracking (MOT) algorithm utilizing location and appearance cues (color and CNN features) to carry out cow tracking process. To leverage the robustness of the system, we also considered the recent features from the previous tracked cow.
-
Dairy Cattle Detection in Loose Housing Calving Pen by Using Semantic Segmentation Networks Reviewed International journal
Swe Zar Maw, Thi Thi Zin, Pyke Tin
ICIC Express Letters, Part B: Applications 13 ( 3 ) 279 - 286 2022.3
Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
When it comes to controlling a cattle farm, being able to accurately forecast when calving will happen can be quite beneficial because it allows employees to assess whether or not assistance is required. If such help is not provided when it is required, the calving process may be prolonged, severely impacting both the mother cow and the calf ’s health. Multiple diseases may result from such a delay. During the production cycle, one of the most crucial events for cows is calving. An accurate video-monitoring technique for cows can spot abnormalities or health issues early, allowing for prompt and effective human interference. To make this surveillance automated, a crucial task is to detect the dairy cattle. For this purpose, in this research, we have proposed an effective semantic segmentation network for segmenting the cow from the 360-degree surveillance camera. The proposed network is a modified version of the U-Net architecture. An additional mod-ule is added in the U-Net architecture which is named as Convolutional Long Short-Term Memory (ConvLSTM) block. The ConvLSTM block allows for effective feature sharing between the less dense layers and denser layers. Experiments with our suggested method were carried out at a big dairy farm in Japan’s Oita Prefecture. The suggested method’s experimental findings demonstrate that it holds promise in real-world applications.
-
Automatic Detection and Tracking of Mounting Behavior in Cattle Using a Deep Learning-Based Instance Segmentation Model Reviewed International coauthorship International journal
Su Myat Noe, Thi Thi Zin, Pyke tin, I. Kobayashi
International Journal of Innovative Computing, Information and Control 18 ( 1 ) 211 - 220 2022.2
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：International Journal of Innovative Computing, Information and Control
In precision livestock farming, estrus detection in cattle is particularly im-portant for cattle breeding management. With accurate estrus detection, artificial in-semination can be administered, which proportionally affects the productivity of livestock farms. Most estrus behaviors can be successfully detected by recognizing the mating postures of cattle. Therefore, in this paper, we propose an estrus detection approach that tracks and identifies cattle mating postures individually based on video inputs. To achieve precise identification and to obtain individual cattle information, segmenting each cattle from its background is a vital step. To solve pixel-level segmentation masks for the cattle in an outer ranch environment, an instance segmentation approach based on a Mask R-CNN deep learning framework is also proposed. In this paper, individual cattle segmentation for detecting the mounting behaviors is carried out first. This is followed by a lightweight tracking algorithm as a post-processing step which is our study innovation. The training data were collected by installing surveillance cameras at a livestock farm, and for the testing data, various datasets from different camera placements were used. The proposed approach achieved 95.5% detection accuracy in identifying the estrus be-haviors of cattle.
-
A Stochastic Modeling Procedure for Predicting the Time of Calving in Cattle Reviewed International journal
Thi Thi Zin, K. Sumi, Pann Thinzar Seint, Pyke Tin, I. Kobayashi, Y. Horii
ICIC Express Letters, Part B: Applications 13 ( 1 ) 49 - 56 2022.1
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
In this paper we introduce a stochastic modeling technique for predicting time to the occurrence of calving events in cattle. Specifically we establish an application procedure of Wald’s fundamental identity in sequential analysis to predict the time to dairy cow calving as accurately as possible. We have well recognized that Wald’s identity is a fairly handy tool for studying the properties of random walks arising in queueing and dam theories and many other stochastics processes. The identity enables us to obtain ab-sorption probabilities of random walks with one or more barriers which can be interpreted as the occurrence of a calving event in cattle. In order to investigate the proposed problem more insight, we consider the activities of a pregnant cow around the calving event as a sequence of random variables forming a random walk. We then derive results for pre-dicted calving times at which an individual cow calving event occurs in a video-monitored maternity barn. For experimentations, two special probability distributions parameterized by using some real-life data are utilized. The outcome results show the proposed method is promising with high accuracy.
-
Image technology based detection of infected shrimp in adverse environments Reviewed International coauthorship International journal
Thi Thi Zin, T. Morimoto, Naraid Suanyuk, T. Itami and Chutima Tantikitti
Songklanakarin Journal of Science and Technology 44 ( 1 ) 112 - 118 2022.1
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Songklanakarin Journal of Science and Technology
In recent years, countries around Japan and especially in Southeast Asia, white spot disease (WSD) is highly infectious and severely damages shrimp aquaculture. At the same time, the various diseases are occurring in shrimp farms. In the early stages of infection, shrimp shows three abnormal behaviors: (1) they appear in the shallow waters of the farm, (2) they do not move and do not eat even when feeding, and (3) they suddenly stop moving. Currently, infected shrimps are found by visual inspection, which places a burden on the farmers and delays the discovery. Therefore, in this paper, we proposed a system for detecting infected shrimp by using image processing technology in order to eliminate the delay of discovery and reduce the burden of farmers. According to our experimental results, the proposed system has 95% precision, 100% recall rate and an accuracy of 96.4% by using hold-out evaluation method.
-
An absorbing markov chain model to predict dairy cow calving time Reviewed International journal
Swe Zar Maw, Thi Thi Zin, Pyke Tin, I. Kobayashi, Y. Horii
Sensors 21 ( 19 ) 2021.10
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Sensors
Abnormal behavioral changes in the regular daily mobility routine of a pregnant dairy cow can be an indicator or early sign to recognize when a calving event is imminent. Image processing technology and statistical approaches can be effectively used to achieve a more accurate result in predicting the time of calving. We hypothesize that data collected using a 360-degree camera to monitor cows before and during calving can be used to establish the daily activities of individual pregnant cows and to detect changes in their routine. In this study, we develop an augmented Markov chain model to predict calving time and better understand associated behavior. The objective of this study is to determine the feasibility of this calving time prediction system by adapting a simple Markov model for use on a typical dairy cow dataset. This augmented absorbing Markov chain model is based on a behavior embedded transient Markov chain model for characterizing cow behavior patterns during the 48 h before calving and to predict the expected time of calving. In developing the model, we started with an embedded four-state Markov chain model, and then augmented that model by adding calving as both a transient state, and an absorbing state. Then, using this model, we derive (1) the probability of calving at 2 h intervals after a reference point, and (2) the expected time of calving, using their motions between the different transient states. Finally, we present some experimental results for the performance of this model on the dairy farm compared with other machine learning techniques, showing that the proposed method is promising.
DOI： 10.3390/s21196490
-
A Study on Diagnosis of Parkinson's Disease by Walking Video Reviewed International journal
T. Haruyama, Thi Thi Zin, K. Sakai, H. Mochizuki
2021 IEEE 10th Global Conference on Consumer Electronics, GCCE 2021 758 - 759 2021.10
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：2021 IEEE 10th Global Conference on Consumer Electronics, GCCE 2021
Parkinson's disease (PD) is a progressive nervous system disorder that accompanied with resting tremor, bradykinesia, muscle rigidity and impaired posture. The diagnosis for gait disturbance in Parkinson's disease is subjective mostly depends on the experience and skills of experts due to lack of quantitative criterion. As a consequence, nonspecialist doctors could easily make wrong assessment for gait disturbance. Therefore, in this paper, we propose a diagnostic method for PD by analyzing the state of walking with the aids of image processing technology. An experiment was conducted using walking videos recording to confirm the effectiveness of the proposed method.
-
Predicting Calving Time of Dairy Cows by Time Series Model
Tunn Cho Lwin, Thi Thi Zin, Yokota Mitsuhiro
宮崎大学工学部紀要 50 87 - 94 2021.9
Language：English Publishing type：Research paper (scientific journal) Publisher：宮崎大学工学部
Calving time prediction is an important factor in dairy farming. The careful monitoring of cows can help to decrease the loss of calf rates during the calving time; moreover, to know the exact time of birth is crucial to make sure timely assistance. However, direct visual observation is time-wasting, and the continuous presence of observers during calving time may disturb cows. Therefore, in this study, the recording from video cameras and counting the number of standing to lying and lying to standing transitions of 25 cows before 72 hours of calving time are used. The time series approaches namely the exponential distribution probability and autoregressive integrated moving average (ARIMA) model are applied to predict the calving time and the root mean square error (RMSE) is used to check the accuracy and error value of the experiment. By these methods, the calving time is predicted with exact time interval by using
exponential probability. Moreover, the ARIMA model is better accuracies in predicting calving time than autoregressive (AR) and moving average (MA) models. -
Real-time action recognition system for elderly people using stereo depth camera Reviewed International journal
Thi Thi Zin, Ye Htet, Akagi Y., Tamura H., Kondo K., Araki S., Chosa E.
Sensors 21 ( 17 ) 2021.9
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Sensors
Smart technologies are necessary for ambient assisted living (AAL) to help family mem-bers, caregivers, and health-care professionals in providing care for elderly people independently. Among these technologies, the current work is proposed as a computer vision-based solution that can monitor the elderly by recognizing actions using a stereo depth camera. In this work, we intro-duce a system that fuses together feature extraction methods from previous works in a novel combination of action recognition. Using depth frame sequences provided by the depth camera, the system localizes people by extracting different regions of interest (ROI) from UV-disparity maps. As for feature vectors, the spatial-temporal features of two action representation maps (depth motion appearance (DMA) and depth motion history (DMH) with a histogram of oriented gradients (HOG) descriptor) are used in combination with the distance-based features, and fused together with the automatic rounding method for action recognition of continuous long frame sequences. The experimental results are tested using random frame sequences from a dataset that was collected at an elder care center, demonstrating that the proposed system can detect various actions in real-time with reasonable recognition rates, regardless of the length of the image sequences.
DOI： 10.3390/s21175895
-
Framework of cow calving monitoring system using video images Reviewed International journal
K. Sumi, Thi Thi Zin, I. Kobayashi, Y. Horii
Journal of Advances in Information Technology 12 ( 3 ) 240 - 245 2021.8
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Journal of Advances in Information Technology
In modern dairy farms, calving is a very critical point in the life cycle of productive cows and has played a major role in making farm profits and welfare of cows. In this time, a tremendous number of researchers have been studied the problem of calving mostly to predict the time about to calve and to investigate calving process by using wearable sensors. Like human beings, cows also have environmental pressures by wearing sensors on their bodies sometimes may cause calving difficulties. Thus in this paper, an automatic video based cow monitoring system is proposed to reduce losses of dairy farms caused from calving problems. Specifically, this paper investigates some behaviors of cows to predict time for calving process including cow movements, tail up, stretching the legs, repeating standing and sitting. In doing so, we focus on increasing movement and tail up. Here, the inter-frame difference is used for analyzing the movement and count in every frame. In addition, by extracting the head and tail position the activity of tail up or not will be recognized so that time for calving can be estimated. Finally, the proposed method for calving is confirmed by using self-collected video sequences.
-
Systematic inclusion study on some rare gemstones of the mogok area, mandalay region, myanmar Reviewed International coauthorship International journal
Htin Lynn Aung, Thaire Phyu Win, Thi Thi Zin
ICIC Express Letters, Part B: Applications 12 ( 8 ) 751 - 756 2021.8
Authorship：Last author Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
The Mogok area is situated in Mogok Township, Mandalay Region. It is bounded by Latitude 22◦ 52′-23◦ 00′ N and Longitude 96◦ 10′-96◦ 33′ E. The rock sequence of the study area consists of medium to high grade metamorphic rocks; marble, gneiss, and intrusive igneous rocks; Kabaing granite, leucogranite and syenite. It is famous for pres-ence of ruby and sapphire. Exceptionally some rare gemstones also are discovered. The present work is especially intended to explain systematically the inclusions of some rare gemstones from the Mogok area. Liquid feather inclusions present in jeremejevite. Two-phase inclusions occur in morganite and petalite. In petalite, tube-like inclusions also present. Opaque inclusion and solid inclusion occur in rutile and treacle granular inclusion and finger print inclusion observe in sinhalite.
-
Feature Detection and Analysis of Cow Motion Classification for Predicting Calving Time Reviewed International journal
Thi Thi Zin, Saw Zay Maung Maung, Pyke Tin and Y. Horii
International Journal of Biomedical Soft Computing and Human Sciences (IJBSCHS) 26 ( 1 ) 11 - 20 2021.7
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (scientific journal)
-
Handwritten character recognition on android for basic education using convolutional neural network Reviewed International journal
Thi Thi Zin, Shin Thant, Moe Zet Pwint, T. Ogino
Electronics (Switzerland) 10 ( 8 ) 2021.4
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Electronics (Switzerland)
An international initiative called Education for All (EFA) aims to create an environment in which everyone in the world can get an education. Especially in developing countries, many children lack access to a quality education. Therefore, we propose an offline self-learning application to learn written English and basic calculation for primary level students. It can also be used as a supplement for teachers to make the learning environment more interactive and interesting. In our proposed system, handwritten characters or words written on tablets were saved as input images. Then, we performed character segmentation by using our proposed character segmentation methods. For the character recognition, the Convolutional Neural Network (CNN) was used for recognizing segmented characters. For building our own dataset, handwritten data were collected from primary level students in developing countries. The network model was trained on a high-end machine to reduce the workload on the Android tablet. Various types of classifiers (digit and special characters, uppercase letters, lowercase letters, etc.) were created in order to reduce the incorrect classification. According to our experimental results, the proposed system achieved 95.6% on the 1000 randomly selected words and 98.7% for each character.
-
Smart irrigation: An intelligent system for growing strawberry plants in different seasons of the year Reviewed International coauthorship International journal
Ye Htet, Htin Kyaw Oo, Thi Thi Zin
ICIC Express Letters, Part B: Applications 12 ( 4 ) 359 - 367 2021.4
Authorship：Last author Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
Agriculture productivity is very important for a country’s economy. There-fore, varying the way of cultivating plants could provide more foods than before and thus smart irrigation would be one of the best solutions. Therefore, the proposed system main-ly focused on strawberry plants to produce fruits in all seasons using intelligent systems within the small-scale farm. The system emphasized automatic drip irrigation and environment adjustment system integrated with sensors to control temperature, water, and fertilizers supply. Moreover, leaf analysis using image processing controlled by the Raspberry Pi is implemented for the detection of plant nutrient deficiency symptoms. As for the communication unit to inform the users via sensors, Internet of Things technology is adopted. The experimental results show that the plants bear fruits efficiently throughout the year by using the proposed irrigation system and also the symptoms can be detected in early stages as soon as they appeared on the leaves.
-
Automatic detection of mounting behavior in cattle using semantic segmentation and classification Reviewed
Su Myat Noe, Thi Thi Zin, Pyke Tin, Ikuo Kobayashi
LifeTech 2021 - 2021 IEEE 3rd Global Conference on Life Sciences and Technologies 227 - 228 2021.3
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：LifeTech 2021 - 2021 IEEE 3rd Global Conference on Life Sciences and Technologies
In cattle farming sector, the accurate detection of estrus plays a vital role because incorrect timing for artificial insemination affects the cattle business. The noticeable sign of estrus is the standing heat, where the cattle standing to be mounted by other cows for a couple of seconds. In this paper, we proposed cattle region detection using deep learning semantic segmentation model and automatic detection of mounting behavior with machine learning classification methods. Based on the conducted experiment, the results show that a mean Intersection of Union (IoU) of 98% on the validation set. The pixel-wise accuracy for two classes (cattle and background) was found to be both 98%, respectively. For the classification, the proposed method compares the four supervised machine learning methods which can detect with the accuracy rate of Support Vector Machine, Naïve Bayes, Logistic Regression and Linear Regression are 87%, 96%, 90%, and 80% respectively. Among them, Naïve Bayes algorithm perform the best. The novelty of this work noticeably implies that deep learning semantic segmentation could be effectively employed as a pre-processing step in segmenting the cattle and background prior to using various classification models.
-
Petrochemical characteristics of the granitoid rocks of Northern Myanmar Reviewed
Htin Lynn Aung, Thaire Phyu Win, Thi Thi Zin
LifeTech 2021 - 2021 IEEE 3rd Global Conference on Life Sciences and Technologies 229 - 230 2021.3
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：LifeTech 2021 - 2021 IEEE 3rd Global Conference on Life Sciences and Technologies
The research area is located on the Mogaung - Kamaing-Hpakant road in Hpakant Township, Kachin State, northern Myanmar. The dominant lithologic units comprise igneous and metamorphic rocks. The present work is mainly intended to establish the petrogenesis of the igneous rocks based on the petrochemical analysis results. The igneous rocks are mainly microgranite and serpentinite. Major element analysis of some rocks was determined by XRF spectrometer and interpreted the genesis of these rock units. On the basis of the petrochemical characteristics, the microgranite of the study area is I-type peraluminous granitoid formed by partial melting of mantle and / or lower crust in the extensional tectonics.
-
Markov chain monte carlo method for the modeling of posture changes prior to calving Reviewed
Thi Thi Zin, Pyke Tin, Pann Thinzar Seint, Yoichiro Horii
LifeTech 2021 - 2021 IEEE 3rd Global Conference on Life Sciences and Technologies 291 - 292 2021.3
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：LifeTech 2021 - 2021 IEEE 3rd Global Conference on Life Sciences and Technologies
An accurate and careful analysis of posture changes for a dairy cow prior to calving plays an important role in making calving time prediction. The patterns of activities such as frequent changes in postures of a pregnant cows during the time closer to calving are utilized as indicators to predict the time of calving. In this paper, we introduce Markov Chain Monte Carlo (MCMC) method to generate the patterns of four states activities such lying, transitions from lying to standing, standing itself and transitions from standing to lying based on the monitored cow activity changes data three days prior to calving. The validity of the generated cow activities in posture changes data is compared with the actual collected data in terms of Euclidean and Cosine distance measures. The experimental results show that the method in this paper can be used as a generalized method to generate synthetic data series of dairy cow activities prior to calving.
-
Activity-integrated hidden markov model to predict calving time Reviewed International journal
K. Sumi, Swe Zar Maw, Thi Thi Zin, Pyke Tin, I. Kobayashi, Y. Horii
Animals 11 ( 2 ) 1 - 12 2021.2
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Animals
Accurately predicting when calving will occur can provide great value in managing a dairy farm since it provides personnel with the ability to determine whether assistance is necessary. Not providing such assistance when necessary could prolong the calving process, negatively affecting the health of both mother cow and calf. Such prolongation could lead to multiple illnesses. Calving is one of the most critical situations for cows during the production cycle. A precise video-monitoring system for cows can provide early detection of difficulties or health problems, and facilitates timely and appropriate human intervention. In this paper, we propose an integrated approach for predicting when calving will occur by combining behavioral activities extracted from recorded video sequences with a Hidden Markov Model. Specifically, two sub-systems comprise our proposed system: (i) Behaviors extraction such as lying, standing, number of changing positions between lying down and standing up, and other significant activities, such as holding up the tail, and turning the head to the side; and, (ii) using an integrated Hidden Markov Model to predict when calving will occur. The experiments using our proposed system were conducted at a large dairy farm in Oita Prefecture in Japan. Experimental results show that the proposed method has promise in practical applications. In particular, we found that the high frequency of posture changes has played a central role in accurately predicting the time of calving.
DOI： 10.3390/ani11020385
-
Consumer behavior analyzer in internet of things (Iot) environments Reviewed
Swe Nwe Nwe Htun, Thi Thi Zin, Pyke Tin
International Journal of Innovative Computing, Information and Control 17 ( 1 ) 345 - 353 2021.2
Language：English Publishing type：Research paper (scientific journal) Publisher：International Journal of Innovative Computing, Information and Control
This paper proposes an analyzer of consumer behavior in Internet of Things (IoT) environments. This analyzer is most useful in predicting the intentions of users during searches, and especially during image searches. Since most technologies are connected on the Internet, search results can be characterized using image-similarity measures. In this paper, information on image similarities is extracted using a Convolutional Neural Network (CNN) in IoT environments. In this proposed consumer behavior analyzer, the similarity measures characterizing the relationships between images are transformed into Markov Chain transition probabilities, and their stationary probabilities are then analyzed to describe the priority order for search results conforming with consumer intentions. In order to confirm the validity of the proposed method, the Yelp public dataset was used. The outcomes using this analyzer are promising, and this analyzer might be instrumental in making further improvements in practical applications of consumer technologies.
-
Image Technology Based Detection of Infected Shrimp in Adverse Environments Reviewed
Thi Thi Zin, Takehiro Morimoto, Naraid Suanyuk, Toshiaki Itami, Chutima Tantikitti
The 1st International Conference on Sustainable Agriculture and Aquaculture: For Well Being and Food Security: Book of Abstracts 115 - 115 2021.1
Language：English Publishing type：Research paper (bulletin of university, research institution)
In recent years, the cultivation of white leg shrimp (Litopenaeus vannamei) has become popular in countries around Japan, especially in Southeast Asia, and at the same time, various diseases have occurred in the farms [1]. In the early stages of infection, shrimp show three abnormal behaviors: (1) they appear in the shallow waters of the farm, (2) they do not move and do not eat even when fed, and (3) they suddenly start moving. Early detection is important step to control this disease because there are no preventive measures. In addition, we are currently visually confirming shrimp that show characteristic of the disease. However, these lead to a burden on the farmers and delay in discovery [2]. Therefore, we propose an image technology based monitoring system for detecting shrimp showing the characteristics of diseases.
-
A study on detecting violence using image processing technology Reviewed
S. Misawa, Thi Thi Zin
ICIC Express Letters, Part B: Applications 12 ( 1 ) 59 - 66 2021.1
Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
In recent years, many security cameras have been installed for crime prevention in downtown areas and public facilities. These cameras have greatly contributed to crime prevention and criminal identification. However, the large number of installed cameras is problematic due to difficulties in manually monitoring and detecting violence and crime in real time, as well as in finding specific video footage recording the inci-dents. This paper describes the use of the background difference method in extracting human regions from data obtained using security cameras. In addition, the paper describes a method of detecting violence using features such as speed and moving distance after contact. Using video footage from seven data sets, these methods have been experimentally evaluated, confirming a high detection rate for incidents involving two people side by side.
-
Intelligent monitoring for elder care using vision-based technology Reviewed
Pann Thinzar Seint, Thi Thi Zin, Pyke Tin
International Journal of Innovative Computing, Information and Control 17 ( 3 ) 905 - 918 2021
Language：English Publishing type：Research paper (scientific journal) Publisher：International Journal of Innovative Computing, Information and Control
Nowadays, smart home care systems are being developed in response to various demands, though challenges remain in realizing various required functionalities. Among many considerations used in developing the proposed system, this paper focuses on ways of recording the consumption of medicine and food by elderly people living alone, as well as ways of communicating information to caregivers. Primarily, we used color coding for objects to facilitate their identification and use. Firstly, we propose useful features, not only between the skin surfaces of hands and mouth, but also the contact between body parts and the objects involved. An Eigen value detector is used to overcome the skin occlusion problem. And then, action detection is performed (such as for picking up or grasping medicine, taking medicine, eating, drinking water, and using a towel) by using a combination of the proposed feature and conditional rule-based learning. Secondly, the proposed system uses context awareness for assessing the subject’s actions using statistical analysis. Finally, the entire system is implemented through the user interface of the application platform. Using this system, caregivers can easily see a record of daily activities, provided with contextual information useful in improving the quality of care. Our proposed system is easy to learn and can provide an economical labor-saving solution for caregivers.
-
A Simple Random Walk Model for Dairy Cow Calving Time Prediction Reviewed
Thi Thi Zin, Pyke Tin, Pann Thinzar Seint, K. Sumi, I. Kobayashi, Y. Horii
2021 IEEE 10th Global Conference on Consumer Electronics, GCCE 2021 756 - 757 2021
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：2021 IEEE 10th Global Conference on Consumer Electronics, GCCE 2021
In this paper, we propose a simple random walk model to predict time of calving event occurs for a pregnant dairy cow. Dairy farmers and experts have well recognized that an accurate calving time prediction is quite important in modern smart dairy farming. To meet these demands, we consider the number of posture changes of a pregnant cow during a few days before the expected dates as a random walk to predict the time at which the calving event occurs. For validation, we show some experimental results by using real life data collected from a large dairy farm in Japan.
-
T. Hayashida, Thi Thi Zin, K. Sakai, H. Mochizuki
2021 IEEE 10th Global Conference on Consumer Electronics, GCCE 2021 760 - 761 2021
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：2021 IEEE 10th Global Conference on Consumer Electronics, GCCE 2021
Discrepancy of exam findings at the same patient make it difficult to ascertain chronological change in the disease and the efficacy of the medicine. Quantitative evaluation of severity is important for improving the discrepancy. In this study, we examine the efficacy of quantitative evaluation of tremor when using single camera. Recording the hand movements of tremor with single camera, and the displacement, velocity, and acceleration signals are acquired using the hand shift between two adjacent video frames. Quantitative evaluation of tremor is performed based on features obtained from each signal. According to the validation results, our method using single camera is possible to classify with an accuracy of up to 82.6%.
-
Cattle Region Extraction using Image Processing Technology Reviewed
Y. Motomura, Thi Thi Zin, Y. Horii
2021 IEEE 10th Global Conference on Consumer Electronics, GCCE 2021 762 - 763 2021
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：2021 IEEE 10th Global Conference on Consumer Electronics, GCCE 2021
In recent years, the number of dairy and beef cattle farms has been decreasing, while the number of cattle and the number of cattle per farm have been increasing, so systems for automatically monitoring cattle have been actively introduced. However, most of them are contact type, which causes physical or mental stress to the cows and is costly when the equipment is damaged. Therefore, in this research, we proposed a method for extracting the approximate shape of cattle using a non-contact 360-degree camera to reduce the burden on livestock farmers and cattle, and confirmed its effectiveness through experiments.
-
Imaging tremor quantification for neurological disease diagnosis Reviewed International journal
Y. Mitsui, Thi Thi Zin, N. Ishii, H. Mochizuki
Sensors (Switzerland) 20 ( 22 ) 1 - 14 2020.11
Authorship：Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Sensors (Switzerland)
In this paper, we introduce a simple method based on image analysis and deep learning that can be used in the objective assessment and measurement of tremors. A tremor is a neurological disorder that causes involuntary and rhythmic movements in a human body part or parts. There are many types of tremors, depending on their amplitude and frequency type. Appropriate treatment is only possible when there is an accurate diagnosis. Thus, a need exists for a technique to analyze tremors. In this paper, we propose a hybrid approach using imaging technology and machine learning techniques for quantification and extraction of the parameters associated with tremors. These extracted parameters are used to classify the tremor for subsequent identification of the disease. In particular, we focus on essential tremor and cerebellar disorders by monitoring the finger–nose–finger test. First of all, test results obtained from both patients and healthy individuals are analyzed using image processing techniques. Next, data were grouped in order to determine classes of typical responses. A machine learning method using a support vector machine is used to perform an unsupervised clustering. Experimental results showed the highest internal evaluation for distribution into three clusters, which could be used to differentiate the responses of healthy subjects, patients with essential tremor and patients with cerebellar disorders.
DOI： 10.3390/s20226684
-
Gemological analysis of some rare gemstones from mogok area, mandalay region Reviewed International coauthorship
Htin Lynn Aung, Thi Thi Zin
ICIC Express Letters, Part B: Applications 11 ( 11 ) 1077 - 1086 2020.11
Authorship：Last author Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
Mogok has long been noted as a supplier of various gemstones over the past decades. The principal gemstones are ruby, sapphire and spinel. Nowadays, fabulous rare gemstones from Mogok are being sold in foreign markets. This area is mainly composed of igneous and metamorphic rocks. Exceptionally rare gemstones are also discovered and they are johachidolite, poudretteite, thorite, etc. The fantastic occurrences of rare gemstones provoke attraction and well attention to mineralogists and gemmologists. Most of the rare gemstones in the present research work are studied from gems dealers from Mogok. Other rare samples are recorded and studied in the favor of the gems collectors. The data on primary occurrence of these rare gemstones are still uncertain and further investigation should be required. In the Mogok area, these rare minerals are recovered from alluvial, eluvial, residual deposits along the riverside, hill slope, flat plains and low-lying area. Economically, rare gemstones are highly important for both local and foreign gem markets. Some gemstones are important economically as well as technologically for its composition, such as thorite and beryl, which are used in space and aeronautical purposes. Most of the rare gemstones are valuable for its rarity and collected as museum pieces and collector’s stones. Thus, they are invaluable.
-
Predicting calving time of dairy cows by exponential smoothing models Reviewed International coauthorship
Tunn Cho Lwin, Thi Thi Zin, Pyke Tin
2020 IEEE 9th Global Conference on Consumer Electronics, GCCE 2020 322 - 323 2020.10
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：2020 IEEE 9th Global Conference on Consumer Electronics, GCCE 2020
In dairy farming, calving time prediction is crucial because the calf loss rate during calving time is rising for many reasons. In this paper, we propose a time series model with exponential smoothing technique to predict the time of calving event occurs. By investigating the changes in behavior patterns a few days before the calving events, the proposed method can predict accurately the time of occurrence of calving events. To be specific, we model the number of changes in behaviors of standing and lying as a time series in hourly basis. We then employ the exponential smoothing techniques and survival probability with exponential distribution to make the prediction process. To confirm the proposed method, some experimental works are performed by using video records of 25 cows in calving pen from a large farm at Oita prefecture, Japan.
-
Image Processing and Statistical Analysis Approach to Predict Calving Time in Dairy Cows Reviewed
Swe Zar Maw, Thi Thi Zin, Pyke Tin
2020 IEEE 9th Global Conference on Consumer Electronics, GCCE 2020 318 - 319 2020.10
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：2020 IEEE 9th Global Conference on Consumer Electronics, GCCE 2020
An accurate prediction of calving time in dairy cows is one of the most important factors to make an optimal reproduction process in dairy farming. This paper proposes an image processing and statistical analysis approach to predict calving time in dairy cows. Specifically, we extract the behavior changes patterns of the expected cows by using simple effective motion history images (MHI) a few days before the occurrence of calving event from the video sequences taken in the maternity bans. We then classify extracted features with support vector machine (SVM) and analyze the behavior changes by using statistical method, Hidden Markov model (HMM) for prediction process. To confirm the validity of proposed method, we perform some experiments by installing 360-degree view cameras at the top of calving bans. At the first stage, we analyzed the behaviors of 25 dairy cows for 72 hours before giving birth. As a result, we find that the proposed method is promising.
-
Feature Detection and Classification of Cow Motion for Predicting Calving time Reviewed
Thi Thi Zin, Saw Zay Maung Maung, Pyke Tin, Y. Horii
2020 IEEE 9th Global Conference on Consumer Electronics, GCCE 2020 305 - 306 2020.10
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：2020 IEEE 9th Global Conference on Consumer Electronics, GCCE 2020
The monitoring and automatic detecting of cow behaviors is a key factor for predicting cow calving times. This paper describes the analysis of cow motion patterns by using 360 camera in order to identify various views of cow states. Firstly, Principle Component Analysis (PCA) is applied to solve the rotation variant problem in different postures of cow body and then the dominant features (shape distances) are extracted for cow motion classification such as Standing, Lying, and Transition States (Standing-to-Lying and Lying-to-Standing). During the movement of cow motions, the increasing and decreasing trends of shape (Beta features) from cow body are used to classify transition activities of cows. We prepared the datasets by grouping similar motion sequences and tested against with the proposed features. According to experimental results, the proposed system can give the high accuracy with low computational cost in case of detecting and classifying cow motions.
-
Elderly monitoring and action recognition system using stereo depth camera Reviewed
Thi Thi Zin, Ye Htet, Y. Akagi, H. Tamura, K. Kondo, S. Araki
2020 IEEE 9th Global Conference on Consumer Electronics, GCCE 2020 316 - 317 2020.10
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：2020 IEEE 9th Global Conference on Consumer Electronics, GCCE 2020
The proposed system used stereo type depth camera by examining the human action recognition and also sleep monitoring in the elderly care center. Different regions of interest (ROI) are extracted using the U-Disparity and V-Disparity maps. The main information used for recognition is 3D human centroid height relative to the floor and percentage of movement from frame differencing for sleep monitoring. The results from the experiments of the proposed method show that this system can detect the person location, sitting or lying and also sleep behaviors effectively.
-
Detection of Estrus in Cattle by using Image Technology and Machine Learning Methods Reviewed
Su Myat Noe,Thi Thi Zin, Pyke Tin, H. Hama
2020 IEEE 9th Global Conference on Consumer Electronics, GCCE 2020 320 - 321 2020.10
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2020 IEEE 9th Global Conference on Consumer Electronics, GCCE 2020
Detection of estrus in cattle in early phase is especially vital in the era of precision farming. This paper focuses on the detection of estrus in cattle by using image processing techniques and machine learning methods. In doing so, we first utilize an image analysis to investigate some behaviors of cattle in estrus, which is standing when mounted by the other cattle. We then extract some statistical measures based on polyline shape features of detected cattle images and utilize these measures as an input to machine learning algorithms. Specifically, in this paper, we employ the three supervised machine learning methods, which is Support Vector Machine (SVM), Logistic Regression (LR), and Multiple Linear Regression (MLR) classifiers. Some experimental works are performed by using real-life video sequences. The results show promising and capable to detect the behavior of estrus both cost-effectively (only image) and specifically with the detection rate of SVM is 97%, LR is 94%, and MLR is 94%, respectively.
-
Dam Water Overflow Estimation using Time Series Reviewed International coauthorship
Mie Mie Khin, Mie Mie Tin, Thi Thi Zin, Pyke Tin
2020 IEEE 9th Global Conference on Consumer Electronics, GCCE 2020 285 - 286 2020.10
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2020 IEEE 9th Global Conference on Consumer Electronics, GCCE 2020
This paper will implement the water level estimation of the dam from Myanmar. We use the time series stochastic model to calculate the water level estimation varying on in flow and consumption of dam. This approach is applying probability of Markovian Time Series. This paper based on rainfall and the other factors of dam water storage such as inflow and outflow of dam. This result estimate actual monthly water spread area and shows error is small. This research result also highlight that Markovian Time Series model is one of the best estimate processes for water level estimation.
-
A mobile application for offline handwritten character recognition Reviewed
Thi Thi Zin, Moe Zet Pwint, Shin Thant
2020 IEEE 9th Global Conference on Consumer Electronics, GCCE 2020 10 - 11 2020.10
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：2020 IEEE 9th Global Conference on Consumer Electronics, GCCE 2020
The handwritten character recognition is a computerized system that is able to identify and recognize characters and words written by user. In this paper, we proposed offline handwritten character recognition using deep learning architecture. The Android application of the proposed system is also created by using OpenCV and TensorFlow Lite. The proposed system is aimed to use as a teaching aid in helping kindergarten to primary school level students, especially for practicing their writing and learning. The local handwritten dataset, which includes digit, English alphabet and mathematical symbols that are collected from students, is used for training and testing operations. According to the experimental results, the proposed system is very promising and it will be a useful application for educational environment.
-
A stochastic model for dairy cow body condition scores changes between two successive calving events Reviewed
Thi Thi Zin, Pyke Tin, H. Hama
ICIC Express Letters 14 ( 10 ) 1009 - 1015 2020.10
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters
In this paper, we shall propose a stochastic model to investigate and analyze the patterns of dairy cow body condition scores between two successive calving events. Also, a robust Markov chain was introduced and used for stochastic evaluation of body condition score fluctuations from time to time. The study confirmed greater beneficial with increased healthy performance, but a great variation among farms needs to be taken into account. The stochastic model can fully describe the pattern and quantify its characteristics composed of the sum of random variables derived from milk yields, feeding intakes and transition periods in the body energy reserves changes. For this purpose, mathematical modeling techniques can be used to develop decision making systems, in order to achieve optimality of dairy farm management systems. In this aspect, the body condition score plays a key role to make the system successfully carried out. That is to achieve maintaining target score in corresponding periods such as a few weeks after calving, early lactation, mid lactation and dry periods. This concept leads to looking into the dairy cow energy reserves problem of within-the-two successive calving events since the body condition score fluctuation is critical especially at the time of calving, with improvements in production. However, a little has known the statistical and probabilistic tools for relating the body condition score pattern change and milk production, feeding management and animal health during the inter-calving periods. Therefore, we shall formulate the problem of energy reserves in dairy cow body, as a stochastic model of special in which inputs (feed intakes), outputs (mike produced) and the body condition score (energy research storage) are used as random variables. Utilizing a generalized gamma distribution and the univariate normal distribution functions for the marginal and joint distributions of the inputs and outputs in the model, the expected change patterns in body condition scores with respect to time are derived and analyzed. In order to confirm the validity of the proposed method, some simulation results are obtained by using the estimated parameters for inputs and outputs derived from real life dataset. These results show that the proposed approach is well suited to analyze the behaviors of dairy cows associations with body condition scores changing patterns.
-
Classification of People’s Emotions during Natural Disasters
Nann Hwan Khun, Thi Thi Zin, Mitsuhiro YOKOTA, Hninn Aye Thant
宮大工学部紀要 49 ( 49 ) 85 - 90 2020.9
Language：English Publishing type：Research paper (bulletin of university, research institution) Publisher：宮崎大学工学部
Identifying the polarity of sentiments expressed by users during disaster events have been widely researched. At a recent time, social media has been successfully used as a proxy to gauge the impacts of disasters in real-time. With the growing of microblog sites on the Web, people have begun to express their opinions and emotions on a wide variety of topics on Twitter and other similar social services. We proposed a visual emotion analysis framework for natural disasters. The proposed framework consists of two components, emotion analysis modeling and geographic visualization. This emotion analysis modeling is mostly targeted in case of determining the emotions of Twitter users pre, peri and post natural disasters to help first responders for better managing the situations such as mental health of survived victims and fund raising after severe natural disasters. This geographic visualization system can help people for better understanding the changes of emotion reactions along with the duration of natural disasters and mostly interested regions of Twitter users on these natural disasters. In this research, the situations in California Fire which is happened in 2018 November is experimented for emotion analysis because the affected people often show their states and emotions via big data social media environment.
-
Recognition Based Segmentation of Handwritten Alphanumeric Characters Entry on Tablet PC
Myat Thiri Wai, Thi Thi Zin, Mitsuhiro YOKOTA, Khin Than Mya
宮大工学部紀要 49 ( 49 ) 79 - 84 2020.9
Language：English Publishing type：Research paper (bulletin of university, research institution) Publisher：宮崎大学工学部
Portable tablet PC are very useful in relevant industry of this age because tablets are elegant in appearance and convenient to use. Important things are noted on tablet by handwriting easily in respective industry. Recognition of handwritten characters automatically on tablet like human’s brain is also necessary to be more convenient. To split each character of different handwritten styles is very difficult and it is the main challenging of handwritten character recognition. The previous handwritten character segmentation approaches are still continuing in different problems because of different handwritten styles. The combination of sliding windows, region of interest (ROI) box and convolutional neural network (CNN) are used to execute recognition based segmentation (implicit) of handwritten characters. This system is intended to perform both segmentation and recognition of tablet based application input handwritten characters. Handwritten data are collected from 24 members of our laboratory using three tablets PC models to perform the experiments.
-
An Image Technology Approach to Dairy Cow Monitoring System Invited
254 - 255 2020.9
Language：Japanese Publishing type：Research paper (conference, symposium, etc.)
DOI： 10.11527/jceeek.2020.0_254
Other Link： https://www.jstage.jst.go.jp/article/jceeek/2020/0/2020_254/_pdf
-
Important object segmentation and tracking using features and ocr Reviewed International coauthorship
Mie Mie Tin, Mie Mie Khin, Nyein Nyein Myo, Thi Thi Zin and Pyke Tin
ICIC Express Letters, Part B: Applications 11 ( 9 ) 855 - 861 2020.9
Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
Image processing and segmentation support to analyze image and big data in many different research sectors. Object tracking in a surveillance video sequence supports to the research areas. This system processes on the security surveillance camera network and extracts the important object track information. First is the extraction of the key frames from videos. These key frames are used to extract objects on different cameras and from different background areas. To extract key frames from surveillance videos, the system uses RGB color values and the block method with diagonal movement. The important object is segmented from one key frame and the system searches that object a crossing of other key frames. To segment an object, the system uses OTSU’s threshold method. The comparison of an important object and the other segmented objects uses color moment features and computes the similarity value based on the histogram. The system handles all key frames to extract the similar objects. To find the tracking region of that object on different background regions, the system uses time information on key frames in the video networks. To extract time information from video key frames, the system uses character extraction and recognition with the Optical Character Recognition (OCR) method on the gray level images.
-
Robust tracking of cattle using super pixels and local graph cut for monitoring systems Reviewed International journal
Y. Hashimoto, H. Hama, Thi Thi Zin
International Journal of Innovative Computing, Information and Control 16 ( 4 ) 1469 - 1475 2020.8
Authorship：Last author Language：English Publishing type：Research paper (scientific journal) Publisher：International Journal of Innovative Computing, Information and Control
This paper proposes a robust tracking method of Japanese black cattle. Development of a cattle monitoring system using non-contact and non-invasive methods to improve productivity is a strong demand from livestock farmers in aged society. As one of elemental technologies to realize it, we focus on tracking of cattle for detecting estrus behaviors using video camera. The conventional methods like inter-frame difference and background subtraction do not work well under supposed environment. So we propose a new updating method of ROI (Region of Interest) and Scribbles (for foreground and background) according to the movement of the centroid of the extracted cattle region. SP (Super Pixel) and LGC (Local Graph Cut) are adopted for robust cattle region extraction. The tracking without updating soon fails before cattle goes out of frame, but the tracking with the proposed updating has been successfully continued until cattle has gone out. Through the experimental results carried at Sumiyoshi Field attached to Miyazaki University, the effectiveness of the proposed method has been confirmed.
-
Body Condition Score Estimation Based on Regression Analysis Using a 3D Camera Reviewed
Thi Thi Zin, Pann Thinzar Seint, Pyke Tin, Yoichiro Horii and Ikuo Kobayashi
sensors 20 ( 13 ) 2020.7
Language：English Publishing type：Research paper (scientific journal)
The Body Condition Score (BCS) for cows indicates their energy reserves, the scoring for which ranges from very thin to overweight. These measurements are especially useful during calving, as well as early lactation. Achieving a correct BCS helps avoid calving difficulties, losses and other health problems. Although BCS can be rated by experts, it is time-consuming and often inconsistent when performed by different experts. Therefore, the aim of our system is to develop a computerized system to reduce inconsistencies and to provide a time-saving solution. In our proposed system, the automatic body condition scoring system is introduced by using a 3D camera, image processing techniques and regression models. The experimental data were collected on a rotary parlor milking station on a large-scale dairy farm in Japan. The system includes an application platform for automatic image selection as a primary step, which was developed for smart monitoring of individual cows on large-scale farms. Moreover, two analytical models are proposed in two regions of interest (ROI) by extracting 3D surface roughness parameters. By applying the extracted parameters in mathematical equations, the BCS is automatically evaluated based on measurements of model accuracy, with one of the two models achieving a mean absolute percentage error (MAPE) of 3.9%, and a mean absolute error (MAE) of 0.13.
DOI： 10.3390/s20133705
-
Automatic cow location tracking system using ear tag visual analysis Reviewed International journal
Thi Thi Zin, Moe Zet Pwint, Pann Thinzar Seint, Shin Thant, S. Misawa, K. Sumi, K. Yoshida
Sensors (Switzerland) 20 ( 12 ) 1 - 18 2020.6
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：Sensors (Switzerland)
Nowadays, for numerous reasons, smart farming systems focus on the use of image processing technologies and 5G communications. In this paper, we propose a tracking system for individual cows using an ear tag visual analysis. By using ear tags, the farmers can track specific data for individual cows such as body condition score, genetic abnormalities, etc. Specifically, a four-digit identification number is used, so that a farm can accommodate up to 9999 cows. In our proposed system, we develop an individual cow tracker to provide effective management with real-time upgrading enforcement. For this purpose, head detection is first carried out to determine the cow’s position in its related camera view. The head detection process incorporates an object detector called You Only Look Once (YOLO) and is then followed by ear tag detection. The steps involved in ear tag recognition are (1) finding the four-digit area, (2) digit segmentation using an image processing technique, and (3) ear tag recognition using a convolutional neural network (CNN) classifier. Finally, a location searching system for an individual cow is established by entering the ID numbers through the application’s user interface. The proposed searching system was confirmed by performing real-time experiments at a feeding station on a farm at Hokkaido prefecture, Japan. In combination with our decision-making process, the proposed system achieved an accuracy of 100% for head detection, and 92.5% for ear tag digit recognition. The results of using our system are very promising in terms of effectiveness.
DOI： 10.3390/s20123564
-
Body Condition Score Assessment of Depth Image using Artificial Neural Network Reviewed International journal
Thi Thi Zin, Pann Thinzar Seint, Pyke Tin, Y. Horii
ACM International Conference Proceeding Series: ICCMS '20: Proceedings of the 12th International Conference on Computer Modeling and Simulation 33 - 37 2020.6
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：ACM International Conference Proceeding Series
Body Condition Score (BCS) is a visual sign for managing feeding program. It can be described by numeric numbers to estimate available fat reserves and it also allows producers to make better management decisions. Ignoring BCS until it becomes too thin or fat may result in production and animal losses economically. In this paper, we proposed the BCS assessment tool by using depth information from 3D camera. The experimental data were collected in Oita prefecture and BCS scores were taken under the guidance of experts. The information of depth images are used as feature vectors to the input of Artificial Neural Network (ANN) classifier. The proposed system has achieved the good results for classifying two groups of BCS (3.5 and 3.75) with the overall accuracy of 86%.
-
Time to Dairy Cow Calving Event Prediction by Using Time Series Analysis Reviewed International journal
Thi Thi Zin, K. Sumi, Pyke Tin
ACM International Conference Proceeding Series: ICCMS '20: Proceedings of the 12th International Conference on Computer Modeling and Simulation 143 - 146 2020.6
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：ACM International Conference Proceeding Series
In these days the precision dairy farming which is utilization of the Information and Communication Technologies (ICT) has become one of front line research topics in dairy science as well as in data science leading to Agriculture 4.0. An increase in on-farm mortality due to the occurrence calving difficulties with late assistance can cause possible problems not only for animal welfare but also economic losses to the farmers. In this aspect, calving is an extremely important event in the life of a dairy cow. On the other hand, the time around calving is also a critical period since clinical disorders, and calving problems can occur. Calving difficulties are also becoming increasingly common with many dairy cows requiring assistance at the time of calving. To maximize welfare and minimize losses due to calving difficulties, all animals need to be individually monitored to identify any calving difficulties or health problems as early as possible. In addition, it is important to know the exact time of calving event occur so that timely assistance can be made. In this paper, we propose a continuous video monitoring system for time-to calving event investigation based on time series analysis to achieve an accurate calving time prediction. In doing so we have employed three time series models of autoregressive, moving average smoothing. At the same time, we have confirmed the validity of the proposed method by using the real life data experimented on the University Dairy Farm and one of large dairy farms in Japan. The experimental results show that the proposed time series method is promising and can lead to a new prospect in modern precision dairy farming.
-
Image processing technique and hidden markov model for an elderly care monitoring system Reviewed
Swe Nwe Nwe Htun, Thi Thi Zin, Pyke Tin
Journal of Imaging 6 ( 6 ) 49 2020.6
Language：English Publishing type：Research paper (scientific journal) Publisher：Journal of Imaging
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/). Advances in image processing technologies have provided more precise views in medical and health care management systems. Among many other topics, this paper focuses on several aspects of video-based monitoring systems for elderly people living independently. Major concerns are patients with chronic diseases and adults with a decline in physical fitness, as well as falling among elderly people, which is a source of life-threatening injuries and a leading cause of death. Therefore, in this paper, we propose a video-vision-based monitoring system using image processing technology and a Hidden Markov Model for differentiating falls from normal states for people. Specifically, the proposed system is composed of four modules: (1) object detection; (2) feature extraction; (3) analysis for differentiating normal states from falls; and (4) a decision-making process using a Hidden Markov Model for sequential states of abnormal and normal. In the object detection module, background and foreground segmentation is performed by applying the Mixture of Gaussians model, and graph cut is applied for foreground refinement. In the feature extraction module, the postures and positions of detected objects are estimated by applying the hybrid features of the virtual grounding point, inclusive of its related area and the aspect ratio of the object. In the analysis module, for differentiating normal, abnormal, or falling states, statistical computations called the moving average and modified difference are conducted, both of which are employed to estimate the points and periods of falls. Then, the local maximum or local minimum and the half width value are determined in the observed modified difference to more precisely estimate the period of a falling state. Finally, the decision-making process is conducted by developing a Hidden Markov Model. The experimental results used the Le2i fall detection dataset, and showed that our proposed system is robust and reliable and has a high detection rate.
-
Usability of Tablet Mobile Devices for Offline Handwritten Character Recognition Reviewed
Thi Thi Zin, T. Otsuzuki
ICIC Express Letters, Part B: Applications 11 ( 6 ) 587 - 593 2020.6
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
In recent years, offline handwritten character recognition has been one of the major topics and challenging research areas in the field of pattern recognition and image processing. Moreover, it is a very problematic research area due to the nature of hand writing styles which can vary from one user to another. On top of that, writings by early learners in formal and non-formal education make the problem more complex and challenging. Generally speaking, the early school age children have diversity in handwriting style, variation in angle, size and shape of characters, making the problems of character recognition more difficult. The number of students has been increasing year by year worldwide. However, due to lack of teachers, many children cannot access high quality education. Therefore, this paper proposes offline character recognition for handwritten characters written on tablet mobile devices. This paper can be considered as an additional supporting system for education of preschool children. This proposed method firstly performs character segmentation process on words acquired from the tablet. In the feature extraction process Histogram of Oriented Gradients (HOG) and Bag of Visual Words (BOVW) are used. Support Vector Machine (SVM) is applied in classification process. Some experimental results are shown to confirm the proposed method.
-
Swe Nwe Nwe Htun, Thi Thi Zin, H. Hama
Applied Sciences (Switzerland) 10 ( 9 ) 3005 2020.5
Language：English Publishing type：Research paper (scientific journal) Publisher：Applied Sciences (Switzerland)
In this paper, an innovative home care video monitoring system for detecting abnormal and normal events is proposed by introducing a virtual grounding point (VGP) concept. To be specific, the proposed system is composed of four main image processing components: (1) visual object detection, (2) feature extraction, (3) abnormal and normal event analysis, and (4) the decision-making process. In the object detection component, background subtraction is first achieved using a specific mixture of Gaussians (MoG) to model the foreground in the form of a low-rank matrix factorization. Then, a theory of graph cut is applied to refine the foreground. In the feature extraction component, the position and posture of the detected person is estimated by using a combination of the virtual grounding point, along with its related centroid, area, and aspect ratios. In analyzing the abnormal and normal events, the moving averages (MA) for the extracted features are calculated. After that, a new curve analysis is computed, specifically using the modified difference (MD). The local maximum (lmax), local minimum (lmin), and half width value (vhw) are determined on the observed curve of the modified difference. In the decision-making component, the support vector machine (SVM) method is applied to detect abnormal and normal events. In addition, a new concept called period detection (PD) is proposed to robustly detect the abnormal events. The experimental results were obtained using the Le2i fall detection dataset to confirm the reliability of the proposed method, and that it achieved a high detection rate.
DOI： 10.3390/app10093005
-
Motion Detection Method for Reducing Foreground Aperture Problem in Background Modelling Reviewed
Thi Thi Zin, Pyke Tin, Cho Nilar Phyo
LifeTech 2020 - 2020 IEEE 2nd Global Conference on Life Sciences and Technologies 260 - 261 2020.3
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：LifeTech 2020 - 2020 IEEE 2nd Global Conference on Life Sciences and Technologies
Motion detection plays as an important role in the implementation of video surveillance system and video analysis applications. The detection of moving foreground objects from the complex background scene is the first step in most of the computer vision based surveillance applications. In this paper, we present a new motion detection method using background modelling technique and moving average feature. For establishing the robust background model, we create the Gamma Mixture Model based on the Gamma distribution function. For handling the foreground aperture problem, we use the feature called moving average which can well recognize the alive silent objects. As post-processing, we handle the shadow removing process by generating the dynamic shadow threshold. The experimental results show that the proposed motion detection method can detect the accurate foreground object even though the foreground object appears as silent background objects in real-time environment.
-
Handwritten Characters Segmentation using Projection Approach Reviewed
Thi Thi Zin, Shin Thant, Ye Htet, Pyke Tin
LifeTech 2020 - 2020 IEEE 2nd Global Conference on Life Sciences and Technologies 107 - 108 2020.3
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：LifeTech 2020 - 2020 IEEE 2nd Global Conference on Life Sciences and Technologies
In the area of optical character recognition, handwritten character segmentation is still an ongoing process. Having good segmentation result can provide the better recognition accuracy. In the proposed system, segmentation is carried out mainly on labelling and projection concepts. The input word is firstly labelled. Then, the modified word is segmented with projection approach. The experiments are performed on local dataset with 1600 words approximately and the system gets segmentation accuracy around 85.75 percentage.
-
Cow Identification System using Ear Tag Recognition Reviewed
Thi Thi Zin, S. Misawa, Moe Zet Pwint, Shin Thant, Pann Thinzar Seint, K. Sumi, K. Yoshida
LifeTech 2020 - 2020 IEEE 2nd Global Conference on Life Sciences and Technologies 65 - 66 2020.3
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：LifeTech 2020 - 2020 IEEE 2nd Global Conference on Life Sciences and Technologies
In precision dairy farming, the valid record of individual cow identification is an important factor in large herds management. In this paper, we propose a cow's ear tag recognition system that can be used in dairy cow management. Firstly, cow's head detection is performed by using You Only Look Once (YOLO) object detector followed by ear tag recognition. The ear tag extraction and recognition processes are carried out by image processing techniques and Convolutional Neural Network (CNN) classifier on detected cow's head images. The experiments are conducted by using videos from dairy farm at Hokkaido prefecture, Japan. The proposed system achieved the reliable results which will support to give the informative status in smart farming.
-
Human Action Analysis Using Virtual Grounding Point and Motion History Reviewed
Swe Nwe Nwe Htun, Thi Thi Zin, H. Hama
LifeTech 2020 - 2020 IEEE 2nd Global Conference on Life Sciences and Technologies 249 - 250 2020.3
Language：English Publishing type：Research paper (international conference proceedings) Publisher：LifeTech 2020 - 2020 IEEE 2nd Global Conference on Life Sciences and Technologies
In this paper, we propose an approach to human action analysis for home care monitoring system in the aspect of image processing and life science technologies. We introduce a new concept of a virtual grounding point representing the position of a target person as an innovated feature for action analysis. Specifically, in developing action analysis, the background subtraction is firstly conducted by applying the Mixture of Gaussian and low rank subspace learning. After that, the graph cut is embedded to enhance the foregrounds in order to detect both of moving and motionless object. Secondly, the virtual grounding point is calculated by using the centroid of silhouette image. Finally, motion of the person is estimated by using timed motion history image in order to improve the accuracy of action analysis. A series of the experiments are conducted to confirm the effectiveness of the proposed method.
-
Some Aspects of Mathematical Modeling Techniques in Dairy Science Reviewed
Thi Thi Zin, Pyke Tin and H. Hama
International Workshop on Frontiers of Computer Vision (IW-FCV) 1 - 8 2020.2
Language：English Publishing type：Research paper (international conference proceedings)
Today world is floating in the ocean of data science which is a multi-disciplinary approach to every aspect of our better life. Agriculture 4.0 is coming in and Industry 4.0 is moving forward so that second Information and Communication revolutionary becomes to play a key role in this era. On the other hands, consumers are demanding more quality dairy products than before. Dairy farmers are eager to use the information and communication technologies that push the precision dairy farming to frontier innovations. Due to a shortage of labor in dairy farms and at the same time the farm sizes are bigger, the utilization of ICT along with artificial intelligence and the internet of things are becoming appropriate technologies. Whatever it may be, fundamentally we need to do some preliminary research prior to large scale and wide applications to the real world. In almost analytical and empirical research, we are trying in the ocean of data science. These data could not be used efficiently unless we do further processing and projections. The raw data need to be processed to obtain useful information and knowledge, which could be used for important on-farms decisions. This kind of process is called mathematical modeling which could support to decision-makers to produce correct and valid decisions. In this paper, we will present and analyze some mathematical models in the frame works of dairy science. Moreover, we shall illustrate how these models could be used in practical ways.
-
Background Modelling Using Temporal Average Filter and Running Gaussian Average Reviewed
Thi Thi Zin, Cho Cho Mar and K. Sumi
International Workshop on Frontiers of Computer Vision (IW-FCV) 1 - 8 2020.2
Language：English Publishing type：Research paper (international conference proceedings)
Object detection is very important and fundamental stage for further post processing stages such as individual identification, action recognition, tracking and behavior analysis. Traditional object detection methods start from back-ground modelling stage. Background modelling is a complex and challenging task which highly depends on the speed of moving objects (foreground objects) and stability of static regions (background region). In this paper, we proposed pixel level background modelling method. In this background modelling task, the two main components of background modelling are proposed; (1) background model initialization by selecting the dynamic frames from video sequence and building the model with Temporal Average filter and (2) updating the back-ground model by using Running Gaussian Average method. These methods are applied to RGB images.
-
The Body Condition Score Indicators for Dairy Cows Using 3D Camera Reviewed
Thi Thi Zin, Pann Thinzar Seint, Pyke Tin, Y. Hori
International Workshop on Frontiers of Computer Vision (IW-FCV) 1 - 8 2020.2
Language：English Publishing type：Research paper (international conference proceedings)
Body Condition Score (BCS) evaluates energy reserves of cows when making nutritional management. This information is also useful in fertility rate and milk production. Keeping cows in appropriate condition throughout the production cycle can improve reproductive efficiency and positive impact for economics. In this paper, we propose the effective indicators for automated BCS measurement system and the experimental data are collected at the milking station. We compute the variation of angular area from automatic anatomical point extraction and learning parameters on Region of Interest (ROI) from 3D information. The statistical analysis of BCS trend on year-round data are presented to give the detailed illustrations for dairy farmers. According to the recorded BCS trends, the proposed method can validate that the lower BCS were obtained in early stage of lactation period rather than other seasons.
-
A Hybrid Visual Stochastic Approach to Dairy Cow Monitoring System Reviewed
Thi Thi Zin, Pyke Tin, I. Kobayashi, H Hama
Transactions on Engineering Technologies 119 - 129 2020.1
Language：English Publishing type：Research paper (scientific journal)
In the era of fourth industrial revolution or Industry 4.0 together with the second information and communication technology era, the way we human beings live, act and do are always challenging by the waves of new technologies such as Internet of Things, Artificial Intelligence, Cloud Computing so on and so forth. These new technologies also drive life science, academic, business and production industries for better sides. Among them, one of challenging and innovative technology in life science industry is precision dairy farming which has been pushed into frontline topic among the academia and industry farm managers. Generally speaking, the precision dairy farming can be defined as the use of technologies advanced or simple to analyze the physical and mental behaviors of an individual cows for specifying health and profitability indicators so that overall management and farm performance are to be improved. In other words, the precision dairy farming will focus on welfare and health care for making the farm returns optimal through the use of technologies. Here, technologies may range from daily milk recording to automatic body conditions scoring for individual cows and an accurate prediction of calving time and reporting unusual occurrences during the delivery times.
In order to realize the objectives of precision dairy farming as a fundamental step, a hybrid visual stochastic approach which is a fusion of image technology and statistical method to dairy cow monitoring system is introduced in this paper. The proposed system will investigate four key areas for the precision dairy farming namely: cow identification, body condition scoring, detection of estrus behavior, prediction of time for the occurrence calving event. In doing so, the combination of image technology, statistical methods and stochastic models will be utilized. Specifically, image processing methods will be performed to detect cow activities such as standing, lying and walking in association with time, space and frequencies. Then collected data are to be transformed into a stochastic model of special type of Markov Chain for decision making process. Then some experimental results will be shown by using both image and statistical data collected from the real life environments and some available datasets. -
Robust tracking of cattle using super pixels and local graph cut for monitoring systems Reviewed
Yukie Hashimoto, Hiromitsu Hama, ThiThi Zin
International Journal of Innovative Computing, Information and Control ６（４） 1469 - 1475 2020
Authorship：Last author Language：English Publishing type：Research paper (scientific journal)
-
A Correlated Random Walk Modeling Method for Dairy Cow Inter-calving Body Condition Score Pattern Analysis Reviewed
ThiThi Zin, Pyke Tin
ICIC Express Letters, 14 2020
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (scientific journal)
-
Robust Tracking of Cattle Using Super Pixels and Local Graph Cut for Monitoring Systems Reviewed
ThiThi Zin
International Journal of Innovative Computing, Information and Control (IJICIC) 16 2020
Publishing type：Research paper (scientific journal)
-
Motion History and Shape Orientation Based Human Action Analysis Reviewed
Swe Nwe Nwe Htun, Thi Thi Zin
2019 IEEE 8th Global Conference on Consumer Electronics, GCCE 2019 754 - 755 2019.10
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：2019 IEEE 8th Global Conference on Consumer Electronics, GCCE 2019
In recent decades, many research works focused on the considerations of falling and post-falling event analysis in the aspect of image processing technology to meet the consumer perspective of users. In this paper, the more informative considerations of human action analysis are developed for the prior state of fall or normal actions. In doing so, the effective background subtraction namely the Mixture of Gaussian with Low-Rank Matrix Factorization model is used to obtain robust and adaptive foreground from the practical video sequences. Then, the spatial-temporal bilateral grid for video sequences is constructed by using a standard graph cut theory to improve the cleaned foreground in order to detect the moving/motionless object region. Then, human actions are analyzed by employing motion history and shape orientation using the approximated ellipse method. The experiments are conducted on publicly available video sequences and our simulated video sequences.
-
Image-Based Feeding Behavior Detection for Dairy Cow Reviewed
K. Shiiya, F. Otsuka, Thi Thi Zin, I. Kobayashi
2019 IEEE 8th Global Conference on Consumer Electronics, GCCE 2019 756 - 757 2019.10
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2019 IEEE 8th Global Conference on Consumer Electronics, GCCE 2019
Feeding behavior is an important source of information to know cow's health status, because it is influenced by the feeding environment, cow's physiological changes and health conditions. However, a cow feeds intermittently throughout the day, it is difficult to measure feeding time by visual measurement at field level. In this paper we propose a measurement method of feeding frequency and feeding time for dairy cow, by detecting the feeding behavior by non-contact using a camera.
-
Handwritten Character Segmentation in Tablet Based Application Reviewed
Myat Thiri Wai, Thi Thi Zin, M. Yokota, Khin Than Mya
2019 IEEE 8th Global Conference on Consumer Electronics, GCCE 2019 760 - 761 2019.10
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2019 IEEE 8th Global Conference on Consumer Electronics, GCCE 2019
Nowadays, modern tablets are widely used in the contribution of universal access to education, equality in the exercise of teaching and learning, aiming for more efficient management and administration. The recognition of handwritten characters on these tablets become a necessary consequence in this technology age. The main challenging of handwritten character recognition is splitting each character of unrestricted handwritten scripts and there is no complete solution yet according to our knowledge. In this system, the combination of sliding windows, Region of Interest (ROI) box and Convolutional Neural Network (CNN) are used to execute implicit segmentation of handwritten characters. For performing the experiments, our own dataset is constructed by collecting handwritten data from 24 members of our laboratory using three different tablet PC models.
-
Feature Analysis for Non Formal Education Project in Myanmar Reviewed
Mie Mie Khin, Mie MIe Tin, Thi Thi Zin, Pyke Tin
2019 IEEE 8th Global Conference on Consumer Electronics, GCCE 2019 890 - 891 2019.10
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2019 IEEE 8th Global Conference on Consumer Electronics, GCCE 2019
In a digitally mediated world, Literacy is vital to live. IT based Digital technology is the creation of quality education. It is the main pillar to develop a sustainable socioeconomic development in the country. Myanmar also starts to develop E-government system for every sector. This paper describes feature selection for non-formal education system using literacy project. With this approach, we consider that feature selection of linear regression the analysis is more suitable than support vector method. We are also continuing analysis of this e-education system for rural area in Mandalay division of Myanmar. This paper analysis based on big data concept.
-
Emotion Analysis of Twitter Users on Natural Disasters Reviewed International coauthorship
Nann Hwan Khun, Thi Thi Zin M. Yokota Hnin Aye Thant
2019 IEEE 8th Global Conference on Consumer Electronics, GCCE 2019 342 - 343 2019.10
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：2019 IEEE 8th Global Conference on Consumer Electronics, GCCE 2019
In this information era, people usually express their views and emotions on a wide range of topics through social networking sites and so the role of emotion analysis in social media has been the subject of considerable research. The idea behind this research is that the emotions people express in their status updates can tell us something about how their emotions fluctuate day-to-day due to natural disasters. In this paper, we targeted for emotion analysis of Twitter users on natural disasters. By identifying these emotions, we can help first responders for better managing the situations such as mental health of survived victims. Our experiment is based on California Camp Fire that is happened in 2018 November.
-
Consumer Technology Perspective of a Trinomial Random Walk Model Reviewed
Thi Thi Zin, Pyke Tin, H. Hama
2019 IEEE 8th Global Conference on Consumer Electronics, GCCE 2019 758 - 759 2019.10
Authorship：Lead author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：2019 IEEE 8th Global Conference on Consumer Electronics, GCCE 2019
Recently, the consumer technology association has widened the horizons of the consumer electronic society to include a variety of technologies such as the Internet of Things, Artificial Intelligence, vertical farming and etc. Moreover, among many others, the consumer technology highlights the importance of organic dairy cows and they are well beings so that timely based energy balance and body condition scores of individual cows are needed to be thoroughly studied. In this paper, we shall introduce a Trinomial Random Walk Model as an illustration of consumer technology for analyzing changes in dairy cow body conditions from time to time. In doing so, we will use the 5-point scale system with increment 0.25. Some experimental simulation results are presented by varying the parameters involved.
-
Image Technology and Matrix-Geometric Method for Automatic Dairy Cow Body Condition Scoring Reviewed
Thi Thi Zin, Pyke Tin, Y. Horii, I. Kobayashi
International Journal of Biomedical Soft Computing and Human Sciences 24 ( 1 ) 2019.7
Language：English Publishing type：Research paper (scientific journal)
Dairy cow body condition scoring is a key and important in nearly every aspect of modern precision dairy farms for management of welfare and healthcare, milk yields and split meals (feeding system), aware heat and reproduction peak, calving in smooth and calf saving in proof so that all in all profitability in use. However, the majority of dairy farmers are not doing the body condition scoring on regular basis due to lack of automation so that it causes time-consuming and very subjective. Thus in this paper, an automatic and efficient dairy cow body scoring system is proposed by using Image Technology and Matrix-Geometric Method. By fusing image technol-ogy and matrix-geometric method as a hybrid can lead to a new and efficient technique of dairy cow body condition scoring system. In order to do so, firstly, the anatomical cow body points are extracted from the back views and top views of cow images. Then the geometric properties of the extracted anatomical points are transformed into a Markov Chain Matrix to determine the body condition scores. For confirmation of the validity of the proposed method, some experimental results are shown by using a public cow body condition database.
-
Complex Human–Object Interactions Analyzer Using a DCNN and SVM Hybrid Approach Reviewed
Cho Nilar Phyo, Thi Thi Zin and Pyke Tin
Applied Sciences (Switzerland) 9 ( 9 ) 2019.5
Language：English Publishing type：Research paper (scientific journal) Publisher：Applied Sciences (Switzerland)
Nowadays, with the emergence of sophisticated electronic devices, human daily activities are becoming more and more complex. On the other hand, research has begun on the use of reliable, cost-effective sensors, patient monitoring systems, and other systems that make daily life more comfortable for the elderly. Moreover, in the field of computer vision, human action recognition (HAR) has drawn much attention as a subject of research because of its potential for numerous cost-effective applications. Although much research has investigated the use of HAR, most has dealt with simple basic actions in a simplified environment; not much work has been done in more complex, real-world environments. Therefore, a need exists for a system that can recognize complex daily activities in a variety of realistic environments. In this paper, we propose a system for recognizing such activities, in which humans interact with various objects, taking into consideration object-oriented activity information, the use of deep convolutional neural networks, and a multi-class support vector machine (multi-class SVM). The experiments are performed on a publicly available cornell activity dataset: CAD-120 which is a dataset of human-object interactions featuring ten high-level daily activities. The outcome results show that the proposed system achieves an accuracy of 93.33%, which is higher than other state-of-the-art methods, and has great potential for applications recognizing complex daily activities.
DOI： 10.3390/app9091869
-
Deep Learning for Recognizing Human Activities Using Motions of Skeletal Joints Reviewed
Cho Nilar Phyo, Thi Thi Zin, Pyke Tin
IEEE Transactions on Consumer Electronics 65 ( 2 ) 243 - 252 2019.5
Language：English Publishing type：Research paper (scientific journal) Publisher：IEEE Transactions on Consumer Electronics
With advances in consumer electronics, demands have increased for greater granularity in differentiating and analyzing human daily activities. Moreover, the potential of machine learning, and especially deep learning, has become apparent as research proceeds in applications, such as monitoring the elderly, and surveillance for detection of suspicious people and objects left in public places. Although some techniques have been developed for human action recognition (HAR) using wearable sensors, these devices can place unnecessary mental and physical discomfort on people, especially children and the elderly. Therefore, research has focused on image-based HAR, placing it on the front line of developments in consumer electronics. This paper proposes an intelligent HAR system which can automatically recognize the human daily activities from depth sensors using human skeleton information, combining the techniques of image processing and deep learning. Moreover, due to low computational cost and high accuracy outcomes, an approach using skeleton information has proven very promising, and can be utilized without any restrictions on environments or domain structures. Therefore, this paper discusses the development of an effective skeleton information-based HAR which can be used as an embedded system. The experiments are performed using two famous public datasets of human daily activities. According to the experimental results, the proposed system outperforms other state-of-the-art methods on both datasets.
-
Multivariate Stochastic Analyzer for Dairy Cow Body Condition Scoring Reviewed
Thi Thi Zin, Kosuke Sumi, Pyke Tin
Digital Image & Signal Processing (DISP'19) 2019.4
Language：English Publishing type：Research paper (international conference proceedings)
In this paper, we introduce a conceptual multivariate stochastic analyzer for assessing body condition scores of individual dairy cows. Specifically, by using digital image technologies and statistical stochastic methods, dairy cow body condition scoring machine is to be established. In modern precision dairy farming, the body condition score (BCS) plays an important role as an indicator for measuring health and wealth of a dairy farm. Based on the BCS, today dairy farm manage systems are improved in in various aspects such as milk production, right time for artificial insemination, prediction calving time and so on. Traditionally, human experts perform visual examinations on the key areas of cow body parts such as hook, pin bones, tail head, short ribs, and backbone starches for scoring. However, well-trained human experts become less and dairy farm sizes are bigger as a consequence manual body condition scoring is almost impractical. Thus, in this paper we propose an image technology based stochastic analyzer for automatic scoring the BCS measures of dairy cows. In order to do so, the proposed analyzer first extracts some key anatomical points of a cow by using two-dimensional images taken from top views. Then, the system will derive some distance and angular features of the anatomical points and employs stochastic sampling techniques for refining the extracted features to produce parameters of multiple regressive prediction models and to assess the body condition scores of all dairy cows in the farm. Finally, to confirm the validity of proposed analyzer, we perform some experiments by a well-known benchmark dataset. The experimental results seem to be promising with an impact of high accuracy.
-
Incorporating Digital Imaging in Dairy Cow Anatomical Features Detection Reviewed
Thi Thi Zin, Cho Nilar Phyo, Pyke Tin
Digital Image & Signal Processing (DISP'19) 2019.4
Language：English Publishing type：Research paper (international conference proceedings)
In today precision dairy farming, the most commonly used technologies include wearable devices which must be attached to cows in one way or another in order to monitor cow’s behaviors. Since the wearable sensors placements may be broken or lost and can burden additional stress to the cows, it is necessary to consider an alternative and effective non-contact monitoring system. For this purpose, digital imaging technologies are suitable due to their capabilities of continuous operation and able to full automation. Thus, in this paper, we propose a digital imaging approach based on topological persistence concepts to precision dairy cow monitoring system focused on automated dairy cow anatomical feature detection. Anatomically these features define as hips, hooks, pin bones, tail-heads and rear regions of the cow body. These features will be utilized in the decision making process if and where a cow is present in an image or video frame. Once the system detects a cow in the image, the system automatically identifies an individual cow. The proposed cow anatomical feature detection and cow identification have the potentials in detecting cow body conditions, health conditions in time, milk production trends and predicting calving time and heat occurrence. Finally, by using videos taken in a real-life cow farm, the experimental results confirm the validity of proposed method.
-
An Algebraic Geometric Imaging Approach for Automatic Dairy Cow Body Condition Scoring System Reviewed
Thi Thi Zin, Pyke Tin, Ikuo Kobayashi and Yoichiro Horii
World Academy of Science, Engineering and Technology (ICPDFTA 2019) 2019.4
Language：English Publishing type：Research paper (conference, symposium, etc.)
Today dairy farm experts and farmers have well recognized the importance of dairy cow Body Condition Score (BCS) since these scores can be used to optimize milk production, managing feeding system and as an indicator for abnormality in health even can be utilized to manage for having healthy calving times and process. In tradition, BCS measures are done by animal experts or trained technicians based on visual observations focusing on pin bones, pin, thurl and hook area, tail heads shapes, hook angles and short and long ribs. Since the traditional technique is very manual and subjective, the results can lead to different scores as well as not cost effective. Thus this paper proposes an algebraic geometric imaging approach for an automatic dairy cow BCS system.
-
Dairy Cow Body Conditions Scoring System Based on Image Geometric Properties Reviewed
Thi Thi Zin, Pyke Tin, Ikuo Kobayashi
2019 IEEE 1st Global Conference on Life Sciences and Technologies (LifeTech 2019) 2019.3
Language：English Publishing type：Research paper (international conference proceedings)
In modernized precision dairy farming, the importance of dairy cow body condition scores is well recognized for making the healthy, wealthy and optimizing milk production. Although a burst amount of researches have been investigated the condition scoring problems from various aspects, not much satisfactory results have been come out yet. So, this paper will propose a geometric imaging approach for an automatic dairy cow body conditions scoring system. Specifically, some significant land marks or anatomical points are to be extracted from the top view image of a cow and their geometrical properties such as angles, length and area are investigated to estimate body condition scores. In doing so, the proposed method will employ techniques of polynomial regression, multiple regression, Markov Chain classification. Finally, some experimental results will be presented by using self-collected datasets and some well-known public datasets. The performance of preliminary results shows promising so that the approach of proposed method can lead to be applicable in real life environments.
-
A Study on Detection of Precursor Behaviors of Estrus in Cattle Using Video Camera Reviewed
Hiromitsu Hama, Tetsuya Hirata, Tsubasa Mizobuchi, Thi Thi Zin
2019 IEEE 1st Global Conference on Life Sciences and Technologies (LifeTech 2019) 2019.3
Language：English Publishing type：Research paper (international conference proceedings)
Development of an estrus detection system by non-contact and non-invasive methods to improve productivity is a strong desire from livestock farmers with aging society. As one of the elemental technologies, we will focus on detecting precursor behaviors of estrus in cattle using video camera. First, we converted from two-dimensional motion on video image to three dimensional one. Next, some features which are well-known as estrus precursor behaviors, were selected, for example, walking speed, trajectory and relative positional relationship of two cattle. Through experimental results, we could confirm the effectiveness of our proposed algorism. As the result, although it is a small case, it was able to detect without any false positive and false negative.
-
OCR Perspectives in Mobile Teaching and Learning for Early School Years in Basic Education Reviewed
Thi Thi Zin, Swe Zar Maw, Pyke Tin
2019 IEEE 1st Global Conference on Life Sciences and Technologies (LifeTech 2019) 2019.3
Language：English Publishing type：Research paper (international conference proceedings)
In these days, teaching and learning systems in schools with the use of mobile or portable devices such as tablets, e-readers, smartphones are becoming keen interests of educators as well as parents and teachers in worldwide. In this aspect, the early years of school children in basic education are the most challenging and important in developing effective and quality education for life. Since they are quite young and unable to dedicate time the need for easy to use and effective learning aids has become vital. Especially their writing skills are somehow needed to be improved with encouragements. In order to do so, the handwritten of characters and numerals performed by the children especially those living in less developing countries should be correctly recognized so that means and ways for remedies and solutions for improvements could be found. Thus the optical character recognition techniques for handwritten alphabets and numerals are moving into front especially the handwritten of children of early years in schools. In this paper, we introduce a Mobile Tutor - an effective and correct way of character segmentation and recognition of messy and unclear handwritten characters to help children learn and practice handwriting and early numeric operations such as addition and subtraction as well as help teachers monitor and review children’s progress. Some experiments are performed by providing tablets to the users and collecting handwritten characters from the users for recognition and analysis.
-
OCR perspectives in mobile teaching and learning for early school years in basic education Reviewed
Thi Thi Zin, Swe Zar Maw, Pyke Tin
2019 IEEE 1st Global Conference on Life Sciences and Technologies, LifeTech 2019 173 - 174 2019.3
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2019 IEEE 1st Global Conference on Life Sciences and Technologies, LifeTech 2019
In these days, teaching and learning systems in schools with the use of mobile or portable devices such as tablets, e-readers, smartphones are becoming keen interests of educators as well as parents and teachers in worldwide. In this aspect, the early years of school children in basic education are the most challenging and important in developing effective and quality education for life. Since they are quite young and unable to dedicate time the need for easy to use and effective learning aids has become vital. Especially their writing skills are somehow needed to be improved with encouragements. In order to do so, the handwritten of characters and numerals performed by the children especially those living in less developing countries should be correctly recognized so that means and ways for remedies and solutions for improvements could be found. Thus the optical character recognition techniques for handwritten alphabets and numerals are moving into front especially the handwritten of children of early years in schools. In this paper, we introduce a Mobile Tutor-an effective and correct way of character segmentation and recognition of messy and unclear handwritten characters to help children learn and practice handwriting and early numeric operations such as addition and subtraction as well as help teachers monitor and review children's progress. Some experiments are performed by providing tablets to the users and collecting handwritten characters from the users for recognition and analysis.
-
Dairy Cow Body Conditions Scoring System Based on Image Geometric Properties Reviewed
Thi Thi Zin, Pyke Tin, I. Kobayashi
2019 IEEE 1st Global Conference on Life Sciences and Technologies, LifeTech 2019 171 - 172 2019.3
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2019 IEEE 1st Global Conference on Life Sciences and Technologies, LifeTech 2019
In modernized precision dairy farming, the importance of dairy cow body condition scores is well recognized for making the healthy, wealthy and optimizing milk production. Although a burst amount of researches have been investigated the condition scoring problems from various aspects, not much satisfactory results have been come out yet. So, this paper will propose a geometric imaging approach for an automatic dairy cow body conditions scoring system. Specifically, some significant land marks or anatomical points are to be extracted from the top view image of a cow and their geometrical properties such as angles, length and area are investigated to estimate body condition scores. In doing so, the proposed method will employ techniques of polynomial regression, multiple regression, Markov Chain classification. Finally, some experimental results will be presented by using self-collected datasets and some well-known public datasets. The performance of preliminary results shows promising so that the approach of proposed method can lead to be applicable in real life environments.
-
A study on detection of precursor behaviors of estrus in cattle using video camera Reviewed
H. Hama, T. Hirata, T. Mizobuchi, Thi Thi Zin
2019 IEEE 1st Global Conference on Life Sciences and Technologies, LifeTech 2019 166 - 167 2019.3
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2019 IEEE 1st Global Conference on Life Sciences and Technologies, LifeTech 2019
Development of an estrus detection system by non-contact and non-invasive methods to improve productivity is a strong desire from livestock farmers with aging society. As one of the elemental technologies, we will focus on detecting precursor behaviors of estrus in cattle using video camera. First, we converted from two-dimensional motion on video image to three dimensional one. Next, some features which are well-known as estrus precursor behaviors, were selected, for example, walking speed, trajectory and relative positional relationship of two cattle. Through experimental results, we could confirm the effectiveness of our proposed algorism. As the result, although it is a small case, it was able to detect without any false positive and false negative.
-
A Hybrid Rolling Skew Histogram-Neural Network Approach to Dairy Cow Identification System Reviewed
Cho Nilar Phyo, Thi Thi Zin, H. Hama, I. Kobayashi
International Conference Image and Vision Computing New Zealand 2018-November 2019.2
Language：English Publishing type：Research paper (international conference proceedings) Publisher：International Conference Image and Vision Computing New Zealand
In this paper, we propose a hybrid method in which rolling skew histogram and neural network techniques are fused to recognize patterns and identify cows in the milking rotary parlor of dairy farms. Individual cow identification is very important for managing the welfare and health care of an individual cow and developing the body condition scoring system. Although there has some sensor-based cows' identification system, those systems require to attach the sensor devices on each cow which is costly and burden on the cows. Since the proposed method applies a single video camera which is a non-contact device for the identification of many different cow patterns, the proposed system is low cost and no burdens on the cow. In particular, the operation of the system takes place while the cows are in the milking process in rotary milking parlor where the monitoring of individual cow is more effective than some other time and places. The identification process is based on the black and white pattern on the cow's body while moving on the rotary milking parlor. For the detecting and cropping of cows' body region is carried out by using rolling the skew histogram and used for the training process of the deep convolutional neural network. The trained network is employed for the identification of individual cow in the testing process. The experiments are performed on the self-collected cow video dataset which includes around 60 different cow's body patterns that have been taken at the large-scale farm in Oita Prefecture, Japan. The experimental results show that the proposed system is promising with the overall accuracy of 96.3 % and it is very effective and practical for the real-time cow identification system needed for establishing a modern precision dairy farming.
-
Framework of Cow Calving Monitoring System Using a Single Depth Camera Reviewed
K. Sumi, Thi Thi Zin, I. Kobayashi, Y. Horii
International Conference Image and Vision Computing New Zealand 2018-November 2019.2
Language：English Publishing type：Research paper (international conference proceedings) Publisher：International Conference Image and Vision Computing New Zealand
Calving difficulty is the primary cause of the problem of increasing death loss in cow-calf. It also profoundly effects on the economic impact of farmers and producers because of calves' death, injury to cows and high veterinary cost. To help the calving difficult of cows in time, long time continuous monitoring is required for deciding when and how to assist the calving process of cows. On the other hand, the continuous monitoring of cows' welfare become the major burden task for labors especially in a large farm where has a large number of cows. Therefore, the demands of sophisticated technology for automatic monitoring of cows' welfare is more and more increasing every year. In recent years, some researcher develops the automatics monitoring and predicting the cows' calving behavior by using the sensors devices such as temperature sensors and acceleration sensors. However, the sensors based system has various problems such as spalling and malfunction of the sensors and even can cause the burden on the cow because sensors are needed to put inside or on the body of the cow. To overcome those problems, in this paper, we propose an automatic detection of cows' calving behavior by using the depth camera (3D camera) along with image processing and computer vision technology and coordinate system transformation concept. The proposed system by using 3D camera can reduce the burden of both labors and cows.
-
A Study on Abnormal Behavior Detection of Infected Shrimp Reviewed
Takehiro Morimoto, Thi Thi Zin, Toshiaki Itami
2018 IEEE 7th Global Conference on Consumer Electronics, GCCE 2018 818 - 819 2018.12
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2018 IEEE 7th Global Conference on Consumer Electronics, GCCE 2018
A method of detecting infected shrimp has been developed. Though shrimp production is thriving in Japan, disease causes much damage. Because the cause is a virus infection, medical treatment is not currently possible. The infection route includes factors such as predation by environmental organisms and water-borne infection. However, no specific countermeasures have been developed. Therefore, early detection of infected shrimp is necessary to prevent secondary infection. When shrimp are infected, they exhibit the following abnormal behaviors: 1) not eating, 2) appearing in shallow water, or 3) making sudden movements. The developed method of detecting infected shrimp involves the use of image processing to determine when the shrimp are not eating.
-
Medication and Meal Intake Monitoring using Human-Object Interaction Reviewed
Pann Thinzar Seint, Thi Thi Zin, Mitsuhiro Yokota
2018 IEEE 7th Global Conference on Consumer Electronics, GCCE 2018 145 - 146 2018.12
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2018 IEEE 7th Global Conference on Consumer Electronics, GCCE 2018
Needs for end-of-life care are rising because of increasing age-related challenges. Among them, maintenance of nutrition and medication play an important role in healthcare of elderly people. In this situation, caregivers need to receive the daily record as a trending of usage which will lead to improvements in the quality of care. Our objective is to establish a video monitoring system for the act of taking medication and eating activity which is designed by using human-object interaction. For the evaluation of medication intake scenario, we propose the hierarchical classification system using hybrid PRNN-SVM model for action classification and activity interpretation. By the contribution of rulebased learning, our system also recognizes drinking/eating activity.
-
Automatic Postmortem Human Identification using Collarbone of X-ray and CT Scan Images Reviewed
Hanni Cho, Thi Thi Zin, N. Shinkawa, R. Nishii, H. Hama
2018 IEEE 7th Global Conference on Consumer Electronics, GCCE 2018 369 - 370 2018.12
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2018 IEEE 7th Global Conference on Consumer Electronics, GCCE 2018
Nowadays, human identification both before and after death is becoming one of the most important issues in various aspects. However, it is not easy to identify human by doctors or forensic experts manually because it consumes much time, especially in large scale victims. Therefore, an automatic human identification system becomes a vital need. For this purpose, we develop a computerized human identification system based on the collarbone of chest X-ray and CT scan images to identify an unknown person after death by using image processing technology.
-
An Automatic Estimation of Dairy Cow Body Condition Score Using Analytic Geometric Image Features Reviewed
Thi Thi Zin, Pyke Tin, I. Kobayashi, Y. Horii
2018 IEEE 7th Global Conference on Consumer Electronics, GCCE 2018 190 - 191 2018.12
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2018 IEEE 7th Global Conference on Consumer Electronics, GCCE 2018
In today modern precision dairy farms, among many dominant factors the Body Condition Score (BCS) has been considered a critical value to optimize milk production, analyzing health problems, insemination timing, and many others. Currently, the BCS is measured by human experts giving time- consuming and varying outcomes from one expert to another so that an automatic estimation system for body condition scoring is needed to be developed. Although there have been some researchers on the topics of the BCS by using image processing techniques, an efficient and satisfactory method has not been found yet. Therefore, in this paper, a new approach to an automatic estimation of dairy cow BCS using analytic geometry image features will be considered. Some experimental results are shown by using the BCS database.
-
A Study on Non-contact and Non-invasive Neonatal Jaundice Detection and Bilirubin Value Prediction Reviewed
Sojiro Kawano, Thi Thi Zin, Yuki Kodama
2018 IEEE 7th Global Conference on Consumer Electronics, GCCE 2018 204 - 205 2018.12
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2018 IEEE 7th Global Conference on Consumer Electronics, GCCE 2018
Neonatal jaundice is a yellowish discoloration of skin and eyes that commonly occurs in newborn babies. It is a physiological phenomenon in neonates and it occurs due to the overproduction of bilirubin, reduction of bilirubin treatment function. The quick and accurate treatment is required for neonatal jaundice because it can lead to nuclear jaundice, cerebral palsy, intellectual disturbance and various sequelae. The investigation methods for examining neonatal jaundice include examination using jaundice meter and blood sampling. However, these methods require continuous monitoring and can cause burden on newborn babies. In this paper, we propose the non-contact and non-invasive detection method for neonatal jaundice using image processing and computer vision technology. The experiments are performed on the data collected by University Hospital, University of Miyazaki. According to the experiments, we confirmed about the usefulness of proposed method which can work effectively for infants.
-
Tsubasa Mizobuchi, Thi Thi Zin, Ikuo Kobayashi, Hiromitsu Hama
2018 IEEE 7th Global Conference on Consumer Electronics, GCCE 2018 815 - 817 2018.12
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2018 IEEE 7th Global Conference on Consumer Electronics, GCCE 2018
Nowadays, the methodology of cattle production has been transferred from natural mating to artificial insemination. Therefore, estrous detection for cattle is very important to determine the time period that farmers conduct artificial insemination. However, the observation of estrous behavior by human's eyes through the whole day become as a burden task when the number of cattle is large. Therefore, most farmers are missing to notice estrous behavior. In this paper, we proposed the method for automated detection of cattle's estrous behavior, which is mounting and standing, by using laser range sensor and image processing technology. Some experiments are carried out at the Sumiyoshi field, Faculty of Agriculture, University of Miyazaki. Through experimental results, the effectiveness of our proposed method was confirmed.
-
Post-Mortem Human Identification Using Chest X-Ray and CT Scan Images Reviewed
Hanni Cho, Thi Thi ZIN, Norihiro SHINKAWA, and Ryuichi NISHII
International Journal of Biomedical Soft Computing and Human Sciences 23 ( 2 ) 51 - 57 2018.12
Language：English Publishing type：Research paper (scientific journal)
In this paper, we develop a computerized human post-mortem identification system by using chest biometrics feature. The main purpose is to identify an unknown person after death. An unknown death body is identified by comparing the chest CT image after death with the X-ray images before death stored in a database to find the highest similarity. The proposed system consists of four main processes: pre-processing, boundary extraction, feature extraction, and similarity calculation and ranking results. All the images are firstly enhanced using Contrast Limited Adaptive Histogram Equalization (CLAHE), and then ribs boundaries are extracted using morphological erosion. After that, the features are extracted using Discrete Fourier Transform (DFT). We use the Euclidean distance to calculate similarity between those features, and then ranking is performed based on the resulted distances. Finally, the system retrieves the ante-mortem X-ray images that are similar to the query image of post-mortem CT image of a death person. Experiments are conducted on dataset collected from the Faculty of Medicine, University of Miyazaki, and our experimental results are compared with the best result of the existing system under the same conditions. From the comparison, our proposed system performs best and gives the accuracy of 74.07%.
DOI： https://doi.org/10.24466/ijbschs.23.2_51
Other Link： https://www.jstage.jst.go.jp/article/ijbschs/23/2/23_51/_pdf/-char/ja
-
Ranking of Influential users based on User-Tweet bipartite graph Reviewed
Radia EL BACHA, Thi Thi Zin
Proceedings of the 2018 IEEE International Conference on Service Operations and Logistics, and Informatics, SOLI 2018 97 - 101 2018.10
Authorship：Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：Proceedings of the 2018 IEEE International Conference on Service Operations and Logistics, and Informatics, SOLI 2018
Today, even if we are still geographically situated, but real-time news and trends are able to reach us no matter how far we are or how huge is the time difference. This is due to Social Networks Services and the strong involvement of people to them through posting, sharing and interacting online with information. Therefore, we propose a study on patterns of information diffusion on Twitter during global scale events. We also introduce a method for identifying influencers and quantify popularity of Tweets on the Network based on a user-tweet bipartite graph.
-
Classification of Shape Images Using K-mean Clustering and Deep Learning Reviewed
Swe Zar Maw, Thi Thi Zin, Mitsuhiro Yokota, Ei Phyo Min
ICIC Express Letter 12 ( 10 ) 1023 - 1023 2018.10
Language：English Publishing type：Research paper (scientific journal)
In this day and age, deep learning becomes attractive method for learning multilevel features and representation of data. In this work, we propose the new preprocessing method using K-mean clustering and the shape image classification using deep learning. Firstly, we handle K-means clustering for image preprocessing because there are many variation in input images. Image preprocessing using K-mean clustering can upgrade the achievement of the accuracy. We apply two dimensional deep convolutional neural network in order to classify 10000 shape images in the BabyAIImageAndQuestion Datasets into three different classes. We trained 10000 shape images that can properly classify near 100% and tested 5000 images that can deliver outstanding performance. The goal of our research is to develop the visual ability of children which includes visual acuity, tracking, color perception, depth perception, and object recognition by effectively applying the deep learning algorithm. We also hope that our proposed method is useful for real world application.
-
A study on detection of abnormal behavior by a surveillance camera image Reviewed
Hiroaki Tsushita, Thi Thi Zin
Advances in Intelligent Systems and Computing 744 284 - 291 2018.6
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Advances in Intelligent Systems and Computing
At present, an enormous amount of accidents and terrorisms has been occurred all over the world not only Japan. Due to the spread of security cameras, the number of occurrences of theft and robbery incidents has been decreasing more and more. Nonetheless, the arrest rate has not improved so much and improvement and rising of the arrest rate are required. The objective of this paper is detection of snatching that involves an event between two persons, and we made an effort to detect snatching in various kinds of situations by using some video scenarios. This video scenarios include the scene of snatching with a bicycle and the scene of non-snatching with normal pedestrian passing. Our proposed methods consist of several steps: background subtraction, pedestrian tracking, feature extraction, and snatch theft detection. We focused on the feature extraction process in details and used weighted decision fusion system based on these parameter, area feature, motion feature, and appearance feature in the paper [1]. We attempted to detect the snatching event from diverse features.
-
A study on violence behavior detection system between two persons Reviewed
Atsuki Kawano, Thi Thi Zin
Advances in Intelligent Systems and Computing 744 302 - 311 2018.6
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：Advances in Intelligent Systems and Computing
Lately, surveillance cameras have been widely used for security concerns to monitor human behavior analysis by using image processing technologies. In order to take into accounts for human rights, costs effectiveness, accuracy of performance the systems so that an automatic—human behavior analytic system shall be developed. –In particular, this paper focused on the action of two person violence and detecting two person fighting each other will be considered. Some experimental results are presented to confirm the proposed method by using ICPR 2010 Contest on Semantic Description of Human Activities (SDHA 2010) dataset.
-
A study on music retrieval system using image processing Reviewed
Emi Takaoka, Thi Thi Zin
Advances in Intelligent Systems and Computing 744 346 - 354 2018.6
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Advances in Intelligent Systems and Computing
The music retrieval system has made it possible to easily search music by simply listening to songs on machine with the spread of smartphone applications. However, it requires enormous data by voice information existing system. Therefore, we proposed new music retrieval system using images visualized by sound signal processing. In the experiment, we used 35 songs and made data for retrieval by extracting a part from them and using only melody information. As a result, the percentage of correct answers that appeared in the top 3 places was 89%, which proved that the proposed method is useful.
-
A study on detection of suspicious persons for intelligent monitoring system Reviewed
Tatsuya Ishikawa, Thi Thi Zin
Advances in Intelligent Systems and Computing 744 292 - 301 2018.6
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Advances in Intelligent Systems and Computing
Currently, surveillance cameras are proliferating for prevention of crimes worldwide and early detection of emergency situations, and they play a very important role in the field of crime prevention and verification against various crimes. With regard to crime recognition and crime arrest, the number of surveillance cameras has been on the rise since it became widespread, and this has also led to crime prevention. However, in most cases, it will be coped after the occurrence of crime, and as for ongoing surveillance for 24 h, the burden on the surveillance side is heavy and there are cases where suspicious people are overlooked. In this paper, by focusing on the action of “Loitering” performed by a criminal using various characteristics of a person, it is possible to automatically determine whether the target person is a “Normal Pedestrian” or “Suspicious Pedestrian”. We will develop algorithms to make it recognizable and confirm the usefulness in terms of crime prevention and crime verification. It is also expected that establishing detection technology will contribute to crime reduction as a deterrent against crime.
-
A survey on influence and information diffusion in twitter using big data analytics Reviewed
Radia El Bacha, Thi Thi Zin
Advances in Intelligent Systems and Computing 744 39 - 47 2018.6
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Advances in Intelligent Systems and Computing
By now, even if we are still geographically situated, we’re able to reach, connect and know about each other through social networks like never before. Among all popular Social Networks, Twitter is considered as the most open social media platform used by celebrities, politicians, journalists and recently attracted a lot of attention among researcher mainly because of its unique potential to reach this large number of diverse people and for its interesting fast-moving timeline where lots of latent information can be mined such as finding influencers or understanding influence diffusion process. This studies have a significant value to various applications, e.g., understanding customer behavior, predicting flu trends, event detection and more. The purpose of this paper is to investigate the most recent research methods related to this topic and to compare them to each other. Finally, we hope that this summarized literature gives directions to other researchers for future studies on this topic.
-
Behavior analysis for nursing home monitoring system Reviewed
Pann Thinzar Seint, Thi Thi Zin
Advances in Intelligent Systems and Computing 744 274 - 283 2018.6
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Advances in Intelligent Systems and Computing
In this paper, we describe the nursing home monitoring system based on computer vision. This system is aimed for an effective and automatic take care of aged persons to monitor appropriate medication intake. Skin region detection for mouth and hand tracking and color label detection for water and medication bottles tracking are mainly performed for initialization. To differentiate the hand and face region, we use the regional properties of head with online learning. Tracking is done by the minimum Eigen values detection. The overlapping area ratios of desired object to body parts are simply used as feature vectors and Pattern Recognition Neural Network is proposed for the decision of simplest action. This paper presents the 7 types of simple actions recognition for medication intake and our experimental results give the promising results.
-
A study on estrus detection of cattle combining video image and sensor information Reviewed
Tetsuya Hirata, Thi Thi Zin, Ikuo Kobayashi, Hiromitsu Hama
Advances in Intelligent Systems and Computing 744 267 - 273 2018.6
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Advances in Intelligent Systems and Computing
In Japan, the detection rate of estrus behavior of cattle has declined from 70% to 55% in about 20 years. Causes include the burden of the monitoring system due to the aging of livestock farmers and oversight of detection of estrus behavior by multiple rearing. Because the time period during which estrus behavior appears conspicuously is nearly the same at day and night, it is necessary to monitor on a 24-h system. In the method proposed in this paper, region extraction of black cattle is performed by combining frame difference and MHI (Motion History Image), then feature detection of count formula is performed using the characteristic and features of the riding behaviors. In addition, as a consideration of the model experiment, a method of detecting the riding behavior by combining the vanishing point of the camera and the height from the foot of the cattle was proposed. The effectiveness of both methods were confirmed through experimental results.
-
Behavior Analysis for Nursing Home Monitoring System Reviewed
Pann Thinzar Seint, Thi Thi Zin
Big Data Analysis and Deep Learning Applications. ICBDL 2018. Advances in Intelligent Systems and Computing 744 274 - 283 2018.6
Language：English Publishing type：Research paper (international conference proceedings)
In this paper, we describe the nursing home monitoring system based on computer vision. This system is aimed for an effective and automatic take care of aged persons to monitor appropriate medication intake. Skin region detection for mouth and hand tracking and color label detection for water and medication bottles tracking are mainly performed for initialization. To differentiate the hand and face region, we use the regional properties of head with online learning. Tracking is done by the minimum Eigen values detection. The overlapping area ratios of desired object to body parts are simply used as feature vectors and Pattern Recognition Neural Network is proposed for the decision of simplest action. This paper presents the 7 types of simple actions recognition for medication intake and our experimental results give the promising results.
-
Image technology based students' feedbacks analyzing system using deep learning Reviewed
Cho Nilar Phyo, Thi Thi Zin, Hiroshi Kamada, Takashi Toriu
Lecture Notes in Engineering and Computer Science 1 330 - 333 2018.3
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Lecture Notes in Engineering and Computer Science
In these days, the integration of technology in teaching-learning process has become a central role in order to redesign a quality education system especially for the development of interactive education. In this concern, technology based analysis on the interaction between students and teacher and the feedback of the students play key roles. Thus, in this paper, we proposed the automatic students' feedbacks analyzing system for the purpose of speeding up the communication between students and teacher in the classroom by using the image processing and deep learning technology. In the proposed system, the students can use the five kind of color cards for answering the questions or for describing their feedbacks. Then the automatic students' feedback analyzing system will analyze the color cards objects by using the camera and describe the analyzed result to the teacher. In this way, the interaction between students and teacher can be faster and can give a lot of benefit for the education system. In order to implement this system, firstly, the color objects segmentation is performed over the input image using the predefined color thresholds. Then, the noise objects are removed by using the predefined maximum size and minimum size thresholds. Finally, the Deep Convolutional Neural Network (DCNN) is applied in order to classify the five color cards objects and non-card color objects. The experiments are performed on the image that have been taken in the large classroom under the different illumination condition. According to the experimental results, the proposed system can robustly analyze the color cards objects with the accuracy of 97.02% on the training data and 94.38% for the testing data. The proposed system can give the ubiquitous (anytime, anywhere) analyzing of the students' feedback in the classroom.
-
Image Technology based Cow Identification System Using Deep Learning Reviewed
Thi Thi Zin, Cho Nilar Phyo, Pyke Tin, Hiromitsu Hama, Ikuo Kobayashi
Lecture Notes in Engineering and Computer Science: Proceedings of The International MultiConference of Engineers and Computer Scientists 2018 320 - 323 2018.3
Language：English Publishing type：Research paper (international conference proceedings)
Today worldwide trending in precision dairy farming is becoming more focus on an individual cow welfare and health rather than group management by using modern technologies including image processing techniques. In such cases, individual cow identification is one of the fundamental ingredients for the success of modern dairy farming. Thus, in this paper we shall explore and examine how image processing technologies can be utilized in analyzing and identifying individual cows along with deep learning techniques. This system is mainly focus on the identification of individual cow based on the black and white pattern of the cow’s body. In our system, firstly we detect the cow’s body which have been placed on the Rotary Milking Parlour by using the inter-frame differencing and horizontal histogram based approach. Then, we crop the cow’s body region by using the predefined distance value. Finally, the cropped images are used as input data for training the deep convolutional neural network for the identification of individual cow’s pattern. The experiments are performed on the self-collected cow video dataset which have been taken at the large-scale ranch in Oita Prefecture, Japan. According the experimental result, our system got the accuracy of 86.8% for automatic cropping of cow’s body region and 97.01% for cow’s pattern identification. The result shows that our system can automatically recognize each individual cow’s pattern very well.
Other Link： http://www.iaeng.org/publication/IMECS2018/
-
A Study on Disease Diagnosis by Tremor Analysis Reviewed
Yuichi Mitsui, Nobuyuki Ishii, Hitoshi Mochizuki, Thi Thi Zin
Lecture Notes in Engineering and Computer Science: Proceedings of The International MultiConference of Engineers and Computer Scientists 2018 1 2018.3
Language：English Publishing type：Research paper (international conference proceedings)
Tremor is a symptom in which a part of the body (hands, feet, head, etc.) trembles involuntarily. Essential tremor and cerebellar disorders are examples of diseases with tremor, but it is sometime difficult to accurately diagnose these diseases by only physical examination, and there is no indicator to quantitatively evaluate the tremor. In this study, we analyzed the Finger-Nose-Finger (FNF) test, which is a physical examination for detecting patients’ tremor, using image processing technology, and proposed an index to discriminate between essential tremor and cerebellar disorders.
Other Link： http://www.iaeng.org/publication/IMECS2018/
-
Markov Chain Techniques for Cow Behavior Analysis in Video-based Monitoring System Reviewed
Thi Thi Zin, Pyke Tin, Hiromitsu Hama, Ikuo Kobayashi
Lecture Notes in Engineering and Computer Science: Proceedings of The International MultiConference of Engineers and Computer Scientists 2018 1 339 - 342 2018.3
Language：English Publishing type：Research paper (international conference proceedings)
In this paper, we shall explore and examine how Markov Chain techniques of stochastic processes can be utilized to analyze cow behaviors in video-based monitoring systems. In this aspect due to shortage of human experts for monitoring the video-based monitoring system has become a powerful technique to replace for human experts. Moreover the image processing methods will play major roles in analyzing visual behaviors such as cow identification, estrus detection, prediction of calving time and body condition scoring etc. Since cow behaviors are changing with respect to times and related to immediate past, Markov Chain models will be very useful enforce the image processing techniques. Although there has been a tremendous amount of methods for detecting estrus, still it needs to improve for achieving a more accurate and practical. Thus in this paper, cow behaviors are to be analyzed by using Gamma Type Markov Chain Models. In particular, image processing methods will be performed to detect cow activities such as standing, lying and walking in association with time, space and frequencies. Then collected data are to be modeled by using Markov Chain for decision making process. As an illustration, we provide some simulation results based on gamma random number generated data.
Other Link： http://www.iaeng.org/publication/IMECS2018/
-
Color and Shape based Method for Detecting and Classifying Card Images Reviewed
Cho Nilar Phyo, Thi Thi Zin, Hiroshi Kamada, Takashi Toriu
Journal of Robotics, Networking and Artificial Life 4 ( 4 ) 287 - 290 2018.3
Language：English Publishing type：Research paper (scientific journal)
This paper proposes an effective method for detecting and classifying card images by using color and shape features. We extract the card color area using color information and remove low possibility regions based on shape feature. Then, we classify the image by taking classroom size and camera distance. In order to confirm the proposed method, we conduct the experiments with our own videos. According to experimental results the proposed method achieves the overall accuracy of 93.93% in various classroom type (small and large).
-
An Innovative Approach to Video Based Monitoring System for Independent Living Elderly People Reviewed
Thi Thi Zin, Pyke Tin, Hiromitsu Hama
Transactions on Engineering Technologies of International MultiConference of Engineers and Computer Scientists (IMECS 2017) 253 - 264 2018.2
Language：English Publishing type：Research paper (scientific journal)
In these days the population of elderly people grows faster and faster and most of them are rather preferred independent living at their homes. Thus a new and better approaches are necessary for improving the life quality of the elderly with the help of modern technology. In this chapter we shall propose a video based monitoring system to analyze the daily activities of elderly people with independent living at their homes. This approach combines data provided by the video cameras with data provided by the multiple environmental data based on the type of activity. Only normal activity or behavior data are used to train the stochastic model. Then decisions are made based on the variations from the model results to detect the abnormal behaviors. Some experimental results are shown to confirm the validity of proposed method in this paper.
-
Use of Computed Tomography and Radiography Imaging in Person Identification Reviewed
Thi Thi Zin, Ryudo Ishigami, Norihiro Shinkawa, Ryuichi Nishii
Transactions on Engineering Technologies of International MultiConference of Engineers and Computer Scientists (IMECS 2017) 311 - 323 2018.2
Language：English Publishing type：Research paper (scientific journal)
After a large scale of a natural or manmade disaster or fatal accident is hit all victims have to be immediately and accurately identified for the sake of relatives or for judicial aspects. Also it is not ethical for human being to lose their identities after death. Therefore, the identification of a person after or before death is a big issue in any society. In most commonly used methods for person identification includes utilization of different biometric modalities such as Finger-print, Iris, Hand-Veins, Dental biometrics etc. to identify humans. However only a little has been known the chest X-Ray biometric which was very powerful method for identification especially during the mass disasters in which most of other biometrics are unidentifiable. Therefore, in this paper, we propose an identification method which utilizes a fusion of computed tomography and radiography imaging processes to identify human body after death based on chest radiograph database taken prior to death. To confirm the validity of the proposed approach we exhibit some experimental results by using real life dataset. The outcomes are more promising than most of existing methods.
-
A study on disease diagnosis by tremor analysis Reviewed
Y. Mitsui, N. Ishii, H. Mochizuki, Thi Thi Zin
Lecture Notes in Engineering and Computer Science 2233 348 - 351 2018
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Lecture Notes in Engineering and Computer Science
Tremor is a symptom in which a part of the body (hands, feet, head, etc.) trembles involuntarily. Essential tremor and cerebellar disorders are examples of diseases with tremor, but it is sometime difficult to accurately diagnose these diseases by only physical examination, and there is no indicator to quantitatively evaluate the tremor. In this study, we analyzed the Finger-Nose-Finger (FNF) test, which is a physical examination for detecting patients' tremor, using image processing technology, and proposed an index to discriminate between essential tremor and cerebellar disorders.
-
A Markov Chain Approach to big data ranking systems Reviewed
Radia El Bacha, Thi Thi Zin
2017 IEEE 6th Global Conference on Consumer Electronics, GCCE 2017 2017-January 1 - 2 2017.12
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2017 IEEE 6th Global Conference on Consumer Electronics, GCCE 2017
In this paper, we propose a Markov Chain Approach to big data ranking systems. In doing so first, we create a transition matrix which will store a calculated score between every applicable pair of nodes on a large scale network. We then compute the stationary distribution of the Markov Chain with the transition matrix and rank the outcomes of this matrix to get the most influential users on the network. We have implemented our algorithm on the neo4j platform and tested it with some sample datasets. The experimental results show that the proposed method is promising for mining influential nodes in big social networks.
-
A study on automatic display system of the archery score for the visually impaired Reviewed
Kazuhisa Shiiya, Thi Thi Zin, Misaki Jomoto, Hitoshi Watanabe
2017 IEEE 6th Global Conference on Consumer Electronics, GCCE 2017 2017-January 1 - 2 2017.12
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2017 IEEE 6th Global Conference on Consumer Electronics, GCCE 2017
In this paper we propose a robust method of displaying the archery score in real time by using image processing techniques to aid for the development of an automated scoring system after archers shoot arrows. Some experimental results are shown by using video sequences taken in an indoor environment.
-
A study on cow monitoring system for calving process Reviewed
Kosuke Sumi, Thi Thi Zin, Ikuo Kobayashi, Yoichiro Horii
2017 IEEE 6th Global Conference on Consumer Electronics, GCCE 2017 2017-January 1 - 2 2017.12
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2017 IEEE 6th Global Conference on Consumer Electronics, GCCE 2017
Calving is a key dairy farming event and the goal is to have deliver a live, healthy calf and mother cow. Thus knowing when a cow is going to calve is very important for a dairy farm since the necessary assistance can be provided in case of difficult calving during and after birth. Thus in this paper we propose an image technology based method for detecting calving process by analyzing the motion features or monitored video sequences. In particular we utilize the motion features of tail up, increasing movement, licking the calf, repeating standing and sitting, stretching the legs. To confirm the proposed method, some experimental results are shown by using video sequences taken in a large dairy farm.
-
Skeleton motion history based human action recognition using deep learning Reviewed
Cho Nilar Phyo, Thi Thi Zin, Pyke Tin
2017 IEEE 6th Global Conference on Consumer Electronics, GCCE 2017 2017-January 1 - 2 2017.12
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2017 IEEE 6th Global Conference on Consumer Electronics, GCCE 2017
Nowadays, deep learning is very popular in a variety of research field due to its outperformance over the existing machine learning methods and its high generality over raw inputs. According to recent surveys, deep learning can give high performance in visual object recognition system. Human Action Recognition (HAR) is a promising research area over the computer vision research field due to its enormous applicability. Most of the conventional HAR need to extract the handcrafted features in advance before classifying the actions and the environments are fixed. Those limitations make HAR to depend too much on the problem. In real-world, it is difficult to choose the suitable feature depending on the problem and difficult to fix the environment. In this paper, we applied the deep learning technology over the Skeleton Motion History Image (Skl MHI) of human actions to implement HAR that can work independently on the problem domain. According to the experimental results, the proposed system achieves the high recognition accuracy with low computational cost under the various environments.
-
Nanyaseik ruby of phakant township, Kachin state, northern Myanmar Reviewed
Htin Lynn Aung, Thi Thi Zin
2017 IEEE 6th Global Conference on Consumer Electronics, GCCE 2017 2017-January 1 - 2 2017.12
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2017 IEEE 6th Global Conference on Consumer Electronics, GCCE 2017
The marble unit of the study area is sub divided into five subunits; depending up on the minerals assemblages: (1) phlogopite marble (2) diopside-chondrodite-marble (3) graphite marble (4) serpentine bearing marble and (5) white marble. Tertiary sedimentary rocks of Pakhren sandstone, equivalent to Namting Series of Chhibber, 1934 occupies the eastempart. Gemstones are found as detrital fragments in gem-bearing soil horizons known as byones. The Nanyaseik rubies are characterized by their distinct colours of which, the commonest colour being, light pinkish red and intense red and rarely pigeon's blood red. In crystal forms, rubies usually have rounded corners rhombohedrons, pinacoids and not well developed prism faces. Habitually, rhombohedral faces display coarse striations and some with pitted surfaces. It is probable that the Nanyaseik area is situated near the plate boundaries and within the northern splay of the Sagaing fault. Moreover, it also forms a segment of Jade Mine region. Therefore, it is reasonable that the pressure had played an important role more effectively than the temperature in the process of metamorphism.
-
Epitome key information extraction using color values on block Reviewed
Mie Mie Tin, Nyein Nyein Myo, Mie Mie Khin, Thi Thi Zin, Pyke Tin
2017 IEEE 6th Global Conference on Consumer Electronics, GCCE 2017 2017-January 1 - 2 2017.12
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2017 IEEE 6th Global Conference on Consumer Electronics, GCCE 2017
In this paper, we propose block diagonal movement technique to classify shots on video frame sequence. Both RGB color pixel values and Hue value are employed in the proposed technique. Since Key frames are essential to analysis on large amount of video sequence databases this paper uses surveillance video files to extract some meaningful information key frames from long video sequences. This purpose is to reduce weak transaction frames in a large video stream by using block base approaches with diagonal movement. The experimental results show that RGB pixel value base approach is more suitable than HSV hue value base approach. In addition, it is learnt that RGB pixel value base method can extract key frames than HSV hue base with less processing times. RGB pixel values base diagonal block method can process accurately, clearly and stability to extract key information frames.
-
Automatic evaluation of Cow's body-condition-score using 3D camera Reviewed
Sosuke Imamura, Thi Thi Zin, Ikuo Kobayashi, Yoichiro Horii
2017 IEEE 6th Global Conference on Consumer Electronics, GCCE 2017 2017-January 1 - 2 2017.12
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2017 IEEE 6th Global Conference on Consumer Electronics, GCCE 2017
Body Condition Scoring (BCS) is a method of evaluating fatness or thinness in cows, and it is important to manage productivity of the cows. However, it is not easy to measure BCS by observing animals because it consumes much time and costs, especially in the large-scale farming. Therefore, almost farmers are not conducting regular evaluation of BCS. In this paper, we propose the noninvasive method for automated evaluation of cow's BCS by using 3D camera and image processing technology.
-
Exploring gemstones in northern part of Myanmar Reviewed
Htin Lynn Aung, Thi Thi Zin
Advances in Intelligent Systems and Computing 579 182 - 188 2017.10
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Advances in Intelligent Systems and Computing
The primary occurrences of gemstones in Nanyaseik area seem to be scarce, the secondary placer gem-bearing deposits are noteworthy. All gemstone occurrences from Nanyaseik area are mainly recovered from secondary deposits (gravels). Gemstones are found as detrital fragments in gem-bearing soil horizons known as byones. According to the drainage characteristics of this area and its environs gem-bearing alluvium had been probably descended from northwestern and western watersheds that created those secondary deposits, especially at the junctions of major streams and their tributaries where local people wash the byone and extract gems. These gems include precious rubies, sapphires (including padparadscha) and others; spinel, tourmaline, zircon, quartz, diopside and almandine garnet.
-
A new conceptual model for big data analysis Reviewed
Thi Thi Zin, Pyke Tin, Hiromitsu Hama
Advances in Intelligent Systems and Computing 579 52 - 58 2017.10
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Advances in Intelligent Systems and Computing
In today modern societies, everywhere has to deal in one way or another with Big Data. Academicians, researchers, industrialists and many others have developed and still developing variety of methods, approaches and solutions for such big in volume, fast in velocity, versatile in variety and value in vicinity known as Big Data problems. However much has to be done concerning with Big Data analysis. Therefore, in this paper we propose a new concept named as Big Data Reservoir which can be interpreted as Ocean in which all most all information is stored, transmitted, communicated and extracted to utilize in our daily life. As a starting point of our proposed new concept, in this paper we shall consider a stochastic model for input/output analysis of Big Data by using Water Storage Reservoir Model in the real world. Specifically, we shall investigate the Big Data information processing in terms of stochastic model in the theory of water storage or dam theory. Finally, we shall present some illustrations with simulation.
-
Some characteristics of nanyaseik area corundum and other assorted gemstones in Myanmar Reviewed International coauthorship
Htin Lynn Aung, Thi Thi Zin
Advances in Intelligent Systems and Computing 579 173 - 181 2017.10
Authorship：Last author, Corresponding author Language：English Publishing type：Research paper (international conference proceedings) Publisher：Advances in Intelligent Systems and Computing
The rock sequence of the study area consists of medium to high grade metamorphic rocks, marble, gneiss and intrusive igneous rocks, mainly biotite microgranite and serpentinite. Although the primary occurrences of gemstones in this area seem to be scarce, the secondary placers gem-bearing deposits are noteworthy. The Nanya rubies are characterized by their distinct colours of which, the commonest colour being, light pinkish red and intense red and rarely pigeon’s blood red. A glassy texture with excellent transparency makes the stone more attractive. In crystal forms, rubies usually have rounded corners, rhombohedrons, pinacoids and not well developed prism faces. Habitually, rhombohedral faces display coarse striations and some with pitted surfaces. It is probable that the Nanyaseik area is situated near the plate boundaries and within the northern splay of the Sagaing fault. Moreover, it also forms a segment of Jade Mine region. Therefore, it is reasonable that the pressure had played an important role more effectively than the temperature in the process of metamorphism.
-
Markov queuing theory approach to internet of things reliability Reviewed
Thi Thi Zin, Pyke Tin, Hiromitsu Hama
Advances in Intelligent Systems and Computing 579 165 - 172 2017.10
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Advances in Intelligent Systems and Computing
In today world a new buzzword Internet of Things has been on the news nearly every day. Some researchers are even using Internet of Every Things. Its potentialities and applicability are now on the cutting edge technology. Also, all most all of business, health care, academic institutions are in one way or another, having to deal with the Internet of Things. So the Internet of Things reliability becomes an important factor. In this paper we proposed a Markov Queuing approach to analyze the Internet of Thing reliability. Since queuing theory investigates the delay and availability of functioning things and Markov concepts take the dependency of Things in the Internet, the combination of these two concepts will make the problem clear and soluble. For illustration, we present some experimental results.
-
Estimating body condition score of cows from images with the newly developed approach Reviewed
Nay Chi Lynn, Zin Mar Kyu, Thi Thi Zin, Ikuo Kobayashi
Proceedings - 18th IEEE/ACIS International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing, SNPD 2017 91 - 94 2017.8
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Proceedings - 18th IEEE/ACIS International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing, SNPD 2017
The Body Condition Score (BCS) is the level of energy reserves in many species, including dairy cattle. For the exact management on dairy farms, the judgment process of BCS is critically important. In this study, the implementation of newly developed approach to estimate body condition score is proposed. Back view images of the cow were used in this system. The area around the tailhead and left and right hooks are segmented automatically and then classified that region for estimating the body condition score. The three main steps conducted are (1) segmentation of cows' images, (2) extraction of region of interest (ROI) by using the convex hull method, and (3) calculation of parameter using moving average method. To confirm this new approach, back view images of various cow types are used and the experimental results confirm its effectiveness with accurate results.
-
An Innovative Deep Machine for Human Behavior Analysis Reviewed
Thi Thi Zin, Pyke Tin, Hiromitsu Hama
Proc. of 12th Intl. Conf. on Innovative Computing, Information and Control (ICICIC2017) 2017.8
Language：English Publishing type：Research paper (international conference proceedings)
In this paper we shall propose an innovative deep machine model to analyze human behaviors without using engineers’ hand crafted features. Due to the tremendous potentials with respect to variety of applications and theoretical point of views the deep machine learning techniques and human behavior analysis have become very important and promising two research areas in these days. We shall be bridging these two research areas to gain more benefits in application wise. In particular, we shall introduce two methods: multilevel neural networks with a series of Gamma Activation Functions and deep convolution neural networks using Markov correlated convolution kernels and variable pooling techniques. In addition, back propagation will be used for parameter optimization in order to minimize the mean square errors. The validity of proposed approach will be confirmed by using available public dataset and giving experimental results for comparisons with the best existing results.
-
Human identification using X-Ray image matching Reviewed
Ryudo ISHIGAMI, Thi Thi Zin, Norihiro SHINKAWA, Ryuichi NISHII
Lecture Notes in Engineering and Computer Science 2227 415 - 418 2017.3
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Lecture Notes in Engineering and Computer Science
Human identification both prior to death and after death is becoming one of the major worldwide issues nowadays for law enforcement aspect as well as social and security aspects. General identification prior to death is possible through comparison of many biometric identifiers. However identification after death is impossible using behavioral biometric such as speech and actions. Moreover, in many circumstances such as natural disasters, air plane crash or a case of identification a couple of weeks later, most of physical biometrics may not be useful for identification due to the decay of some soft tissues. A lot of research has been done in the field of different biometric modalities like Finger-print, Iris, Hand-Veins, Dental biometrics etc. to identify humans. However only a little has been known the chest X-Ray biometric which was very powerful method for identification especially during the mass disasters in which most of other biometrics are unidentifiable. Therefore in this paper, we propose a stochastic modelling approach for human identification after death by using chest X- Ray prior to death database. Some experimental results are shown based on real life dataset and confirmed.
-
Visual Monitoring System for Elderly People Daily Living Activity Analysis Reviewed
Thi Thi Zin, Pyke Tin, Hiromitsu Hama
Lecture Notes in Engineering and Computer Science 2227 140 - 142 2017.3
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Lecture Notes in Engineering and Computer Science
In today modern world, due to the increasing number of older and vulnerable people better and new approaches are needed to support people in their own homes. Especially, intelligent visual monitoring system recognizes the activities of elderly people daily living become more and more important. This paper proposes a new approach for detecting the daily activities of elderly people whether they are deviated from normal behavior. In this context we will focus on building stochastic models of behavior based on types of activity. Models are trained using only normal behavior. Variations from the models are considered as abnormal behaviors and these can be highlighted for subsequent review or intervention. Experimental results are shown by using some real life datasets to illustrate the proposed models.
-
An effective method for detecting snatch thieves in video surveillance Reviewed
Hiroaki Tsushita, Thi Thi Zin
Proceeding of The 2017 International Conference on Artificial Life and Robotics (ICAROB 2017) 303 - 306 2017.1
Language：English Publishing type：Research paper (international conference proceedings)
Other Link： http://alife-robotics.co.jp/members2017/icarob/data/html/data/OS_pdf/OS20/OS20-1.pdf
-
Color and Shape based Method for Detecting and Classifying Card Images Reviewed
Cho Nilar Phyo, Thi Thi Zin, Hiroshi Kamada, Takashi Toriu
Proceeding of The 2017 International Conference on Artificial Life and Robotics (ICAROB 2017) 307 - 310 2017.1
Language：English Publishing type：Research paper (international conference proceedings)
Other Link： http://alife-robotics.co.jp/members2017/icarob/data/html/data/OS_pdf/OS20/OS20-2.pdf
-
Automatic Assessing Body Condition Score from Digital Images by Active Shape Model and Multiple Regression Technique Reviewed
Nay Chi Lynn, Thi Thi Zin, Ikuo Kobayashi
Proceeding of The 2017 International Conference on Artificial Life and Robotics (ICAROB 2017) 311 - 314 2017.1
Language：English Publishing type：Research paper (international conference proceedings)
Other Link： http://alife-robotics.co.jp/members2017/icarob/data/html/data/OS_pdf/OS20/OS20-3.pdf
-
Block Based Approach for Key Frame Extraction on Large Video Sequences Reviewed
Mie Mie Tin, Nyein Nyein Myo, Mie Mie Khin, Thi Thi Zin
ICIC express letters. Part B, Applications : an international journal of research and surveys 7 ( 12 ) 2713 - 2717 2016.12
Language：English Publishing type：Research paper (scientific journal)
Other Link： http://www.ijicic.org/elb-7(12).htm
-
A study of cow detection and extraction using feature of contrast rate Reviewed
Kosuke Sumi, Thi Thi Zin, Thu Zar Tint, Tin Myint Naing
International Journal of Research in Advanced Engineering and Technology 2 ( 6 ) 5 - 7 2016.11
Language：English Publishing type：Research paper (scientific journal)
In this day, people who become livestock farmers that workload is big are decreasing. When the cow especially acts the calving behavior, livestock farmers need to watch it for a long time. Therefore, the research is aimed for developing cow calving monitoring system to reduce burden of livestock. In the cow calving monitoring system, cow detection and extraction play a vital role. In this paper, the cow detection and extraction are focused on. In order to implement it, image obtained from video camera was analyzed and identified by using techniques of image processing. In the paper, black-haired cows from video sequences are detected and extracted based on inter-frame difference method and contrast rate features. Sobel edge detector and morphological operation are employed to complete results. The experimental results show strong points and weak points of the system.
Other Link： http://www.newengineeringjournal.in/archives/2016/vol2/issue6
-
Deep Learning Model for Integration of Clustering with Ranking in Social Networks Reviewed
Thi Thi Zin, Pyke Tin, Hiromitsu Hama
Genetic and Evolutionary Computing. ICGEC 2016. Advances in Intelligent Systems and Computing, vol 536. Springer 536 247 - 254 2016.10
Language：English Publishing type：Research paper (international conference proceedings)
-
Automatic Target Tracking System based on Local and Global Features Reviewed
Thi Thi Zin, Kenshiro Yamada
Genetic and Evolutionary Computing. ICGEC 2016. Advances in Intelligent Systems and Computing, vol 536. Springer 536 255 - 262 2016.10
Language：English Publishing type：Research paper (international conference proceedings)
-
Reliability and Availability Measures for Internet of Things Consumer World Perspectives Reviewed
Thi Thi Zin, H. Hama, Pyke Tin
Proc. of The 5th 2016 IEEE Global Conf. on Consumer Electronics (GCCE2016) 367 - 368 2016.10
Language：English Publishing type：Research paper (international conference proceedings)
DOI： 10.1109/GCCE.2016.7800446
Other Link： http://ieeexplore.ieee.org/document/7800446/
-
User-Intent Visual Information Ranking System Reviewed
Swe Nwe Nwe Htun, Thi Thi Zin, M. Yokota, Khin Mo Mo Tun
Proc. of The 5th 2016 IEEE Global Conf. on Consumer Electronics (GCCE2016) 15 - 16 2016.10
Language：English Publishing type：Research paper (international conference proceedings)
DOI： 10.1109/GCCE.2016.7800316
Other Link： http://ieeexplore.ieee.org/document/7800316/
-
Shape descriptor for binary image retrieval Reviewed
Moe Zet Pwint; Thi Thi Zin; Mitsuhiro Yokota; Mie Mie Tin
Proc. of The 5th 2016 IEEE Global Conf. on Consumer Electronics (GCCE2016) 365 - 366 2016.10
Language：English Publishing type：Research paper (international conference proceedings)
DOI： 10.1109/GCCE.2016.7800445
Other Link： http://ieeexplore.ieee.org/document/7800445/
-
Markov Chain based Query Classification for User Intent Image Search Engines Reviewed
Thi Thi Zin, Pyke Tin and H. Hama
ICIC Express Letters 10 ( 9 ) 2129 - 2134 2016.9
Language：English Publishing type：Research paper (scientific journal)
-
Markov chain based query classification for user intent image search engines
Zin T., Tin P., Hama H.
ICIC Express Letters 10 ( 9 ) 2129 - 2134 2016.9
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters
© 2016 ICIC International. Today Web image search engines have well recognized that a challenging and difficult problem of determining the user intent of Web searches should be thoroughly investigated. In this aspect we propose a Markov chain query classification method to understand user behavior and intent based on the queries which users used for searches. Specifically, we analyze some samples of queries from different Web search engines and classify the user queries into three classes such as informational, navigational and transactional classes. By assuming these three classes represent the type of contents that a user desired to express his or her intents, we implement the Markov chain based classification process. Some experimental results are shown to capture user intents by using a set of queries submitted to the Web search engines.
-
The Identification of Dairy Cows Using Image Processing Techniques Reviewed
Thi Thi Zin, S. Sakurai, K. Sumi, I. Kobayashi, H. Hama
ICIC Express Letters, Part B: Application 7 ( 8 ) 1857 - 1862 2016.8
Language：English Publishing type：Research paper (scientific journal)
-
The identification of dairy cows using image processing techniques
Zin T., Sakurai S., Sumi K., Kobayashi I., Hama H.
ICIC Express Letters, Part B: Applications 7 ( 8 ) 1857 - 1862 2016.8
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
© 2016 ICIC International. Systems to recognize and identify individual cows have become more and more important for modern dairy farms to make precise and successful dairy management. Due to this, various types of identification systems have been coming to front line literature. Most of cow identification systems utilize electrical identification methods based on the RFID (radio frequency identification). However, RFID implementations can cause several problems. Therefore, in this paper we propose an effective and easy-to- use cow identification system by using image processing techniques. Specifically, the proposed system involves recognition of individual cow by identifying the landmark patterns on the cow body along with background modeling, and foreground marks extraction tuned on morphological operations. In order to confirm the system developed in this paper, we present some experimental results based on self-collected video sequences taken at Sumiyoshi Field Science Center, the University of Miyazaki.
-
Action Recognition System with the Microsoft KinectV2 using a Hidden Markov Model Reviewed
M. Fujino, Thi Thi Zin
Proc. of The Third Intl. Conf. on Computing Measurement Control and Sensor Network (CMCSN-2016) 118 - 121 2016.5
Language：English Publishing type：Research paper (international conference proceedings)
Other Link： http://ieeexplore.ieee.org/document/8008654/
-
A General Video Surveillance Framework for Cow Behavior Analysis Reviewed
Thi Thi Zin, I. Kobayashi, Pyke Tin, H. Hama
Proc. of The Third Intl. Conf. on Computing Measurement Control and Sensor Network (CMCSN-2016) 130 - 133 2016.5
Language：English Publishing type：Research paper (international conference proceedings)
Other Link： http://ieeexplore.ieee.org/document/8008657/
-
Estrus Detection for Dairy Cow Using a Laser Range Sensor Reviewed
Thi Thi Zin, H. Kai, K. Sumi, I. Kobayashi, H. Hama
Proc. of The Third Intl. Conf. on Computing Measurement Control and Sensor Network (CMCSN-2016) 162 - 165 2016.5
Language：English Publishing type：Research paper (international conference proceedings)
Other Link： http://ieeexplore.ieee.org/document/8008665/
-
Visual analysis framework for two-person interaction
Zin T., Kurohane J.
2015 IEEE 4th Global Conference on Consumer Electronics, GCCE 2015 519 - 520 2016.2
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：2015 IEEE 4th Global Conference on Consumer Electronics, GCCE 2015
© 2015 IEEE. In this paper, a novel approach to two person interaction method is presented in which pose representation is based on the feature of silhouette images. Today human activity recognition and analysis has a tremendous potential to impact a wide range of applications from surveillance to human computer interfaces to content based video retrieval. Specifically, the proposed method makes use of human silhouettes to classify actions and interactions of human present in a scene video. The classes of interactions will include punching, pushing, kicking, hand-shaking, and hugging. Moreover, the detected interactions are further divided into violence or non-violence so that a suitable security measures would be taken. To confirm the validity of the proposed method, the experimental results are carried out by using the publicly available dataset.
-
Block based approach for key frame extraction on large video sequences
Tin M., Myo N., Khin M., Zin T.
ICIC Express Letters, Part B: Applications 7 ( 12 ) 2713 - 2717 2016
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
© 2016 ISSN 2185-2766. Video segmentation and key frame extraction are very important in real world video systems. Key frames are essential to analyze on large amount of video frame sequences. This paper emphasizes surveillance video and the aim is to extract some meaningful key frames from long video sequences. This purpose is to reduce weak transaction frames in a large video stream by using block base algorithm and cluster the transition video shots. Kullback-Leible divergence method is used in key frame extraction for strong transition video shot. For weak transition video shot, the system will find three candidate key frames and they are compared. Key frames are meaningful frames for video sequences. These frames which represent video streams can be analyzed. Duplicated key frames from the video stream are analyzed in order to be extracted from different shots. Finally, key frames have many assets such as stability, accuracy, and summarize information for a large video.
-
A novel research topic ranking system in academic networks
Zin T., Tin P., Hama H.
Advances in Intelligent Systems and Computing 387 361 - 368 2016
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Advances in Intelligent Systems and Computing
© Springer International Publishing Switzerland 2016. In today world, various types of communication networks such as academic, social, technological, business and etc. come into front. All of the networks are continuously growing and expanding in volume, velocity and variety like popular platform Big Data. Among them, the academic networks such as Alumni, Research Gate, Student Network, Teacher Network and so on provide a powerful abstraction of the academic structure and dynamics of diverse kinds of inter personal academic activities and interaction. Generally, the academic network contents such as research findings and educational concepts are created and consumed by the influences of all different academic navigation paths that lead to the challenging research issues. Therefore, identifying important and researcher relevant refined structures such as new research topics information or academic communities become major factors in modern decision making world.. In this paper, we propose a novel research topic ranking system in academic networks by using the research data relational graphs from academic media platform jointly with educational data to improve the relevance between research topics and researchers intentions (i.e., academic relevance). Specifically, we propose a stochastic model based Academic-Research Topic Ranking algorithm by taking academic value into account. By using some extensive experiments, we demonstrated the significant and effectiveness of the proposed academic-research topic ranking method.
-
Zin T., Lin J., Pan J., Tin P., Yokota M.
Advances in Intelligent Systems and Computing 387 2016
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Advances in Intelligent Systems and Computing
-
A novel method for product brand ranking in consumer networks Reviewed
Thi Thi Zin, Pyke Tin, H. Hama
Proceedings of The 4th IEEE Global Conf. on Consumer Electronics (GCCE 2015) 328 - 329 2015.10
Language：English Publishing type：Research paper (international conference proceedings)
-
Visual analysis framework for two-person interaction Reviewed
Thi Thi Zin, J. Kurohane
Proceedings of The 4th IEEE Global Conf. on Consumer Electronics (GCCE 2015) 519 - 520 2015.10
Language：English Publishing type：Research paper (international conference proceedings)
-
A color constancy model for non-uniform illumination based on correlation matrix Reviewed
T. Toriu, M. Hironaga, H. Kamada, Thi Thi Zin
Proceedings of the Tenth International Multi-Conference on Computing in the Global Information Technology 40 - 46 2015.10
Language：English Publishing type：Research paper (international conference proceedings)
-
Cow identification by using shape information of pointed pattern
K. Sumi, I. Kobayashi, Thi Thi Zin
Genetic and Evolutionary Computing (Proceedings of the 9th Intl. Conf. on Genetic and Evolutionary Computing) 2 273 - 280 2015.8
Language：English Publishing type：Research paper (scientific journal)
-
A novel research topic ranking system in academic networks Reviewed
Thi Thi Zin, Pyke Tin, H. Hama
Genetic and Evolutionary Computing (Proceedings of the 9th Intl. Conf. on Genetic and Evolutionary Computing) 1 361 - 368 2015.8
Language：English Publishing type：Research paper (international conference proceedings)
-
A Matrix-Geometric Method for Web Page Ranking Systems Reviewed
Thi Thi Zin, Pyke Tin, H. Hama, T. Toriu
Journal of Information Hiding and Multimedia Signal Processing 6 ( 4 ) 639 - 647 2015.7
Language：English Publishing type：Research paper (scientific journal)
-
A Triplet Markov Chain Model for Loitering Behavior Detection Reviewed
Thi Thi Zin, Pyke Tin, H. Hama, T. Toriu
ICIC International, ICIC Express Letters (Part B: Applications) 6 ( 3 ) 613 - 618 2015.3
Language：English Publishing type：Research paper (scientific journal)
-
A Novel Hybrid Approach to Image Ranking System Reviewed
Pyke Tin, Thi Thi Zin, T. Toriu and H. Hama
ICIC International, ICIC Express Letters (Part B: Applications) 6 ( 3 ) 743 - 748 2015.3
Language：English Publishing type：Research paper (scientific journal)
-
Image smoothing using a metric tensor for an affine invariant scale space
Toriu T., Zin T., Hama H.
2014 4th International Conference on Image Processing Theory, Tools and Applications, IPTA 2014 2015.1
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：2014 4th International Conference on Image Processing Theory, Tools and Applications, IPTA 2014
© 2014 IEEE. This paper proposes a new image smoothing method using a metric tensor for affine invariant scale space. In the field of image processing and recognition, Gaussian filtering is a common procedure for image smoothing. For example, scale space construction based on Gaussian filtering is sometimes used as a preprocessing of various image processing tasks. However Gaussian filtering is not affine invariant. This paper proposes a new method for image smoothing that is invariant under such affine transformation that does not change the area of any region in the image. It is shown that a scale space representation can be constructed collaterally with the image smoothing. Experimental results show that the proposed method is almost never affected by affine transformation different from usual Gaussian filtering. In the proposed method, processing results are expected to be not affected much by variation of the viewpoint.
-
Cow identification by using shape information of pointed pattern Reviewed
K. Sumi, I. Kobayashi, Thi Thi Zin
Advances in Intelligent Systems and Computing 388 273 - 280 2015
Language：English Publishing type：Research paper (scientific journal) Publisher：Advances in Intelligent Systems and Computing
Monitoring cow behavior and performance plays an important role in dairy health and welfare management systems. Due to the increased number of elderly farm workers, the demands for automatic cow monitoring system become a key role. Moreover, the image based technology for a monitoring system is a promising technique because it is relatively low cost and easy to install. In this aspect, the fundamental and important work to be done is to make an identification of individual cows with high accuracy. Thus in this paper, a simple and effective method for cow identification is introduced by using a modified background subtraction method and histogram based decision process. Specifically, the painted-marks are placed on all black-haired cows and video images are taken. Then the marked region is extracted by using the proposed background subtraction method and histogram based features. Finally, the identification process is performed and some experimental results are shown by using self-collected database taken in the University dairy farm.
-
A triplet Markov chain model for loitering behavior detection
Zin T., Tin P., Hama H., Toriu T.
ICIC Express Letters, Part B: Applications 6 ( 3 ) 613 - 618 2015
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
© 2015 ICIC International. Currently, a sizable number of monitoring systems have been established in public areas where the security factors are sensitive to discourage and prevent crimes. One of applications in such places is to detect loitering people since they can lead to various forms of criminal actions such as drug dealing, pick-pocking, theft and many others. To address this problem, we propose a new concept of Triplet Markov Chain model for detecting loitering people in public areas. After having extracted the foreground object, people are recognized amongst the foreground objects. Then they start being embedded into a Markov Chain by using three types of processes, namely, observing process, state process and underlying process. Finally, if the targeted state remains in the scene longer than the user-defined loitering time threshold then the loitering behavior is detected. The proposed method has been confirmed by using our own video sequences and PETS 2007 datasets.
-
A novel hybrid approach to image ranking system Reviewed
Pyke Tin, Thi Thi Zin, T. Toriu, H. Hama
ICIC Express Letters, Part B: Applications 6 ( 3 ) 743 - 748 2015
Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
Today image ranking systems mostly aim at bridging visual images and user intentions through content-based image retrieval concepts. However, it is still far from satisfactory to narrow down the gap between the outcome images and users expectations. Many researchers recognized that by using visual content or textual content alone it is not easy to identify the user’s interest. In this paper, we propose a novel hybrid approach to image re-ranking system which integrates both visual information of an image and user concepts by forming image interest groups. Specifically, we first establish two stochastic matrices namely, image-content matrix and user-intent matrix, and then the two matrices are combined as a convex combination to form a new user-intent-visual matrix. Finally, the stationary distribution of the matrix is used as an image ranking system so that the users can obtain the desired images from the image datasets easily and effectively. Some experimental results are given to confirm our proposed method.
-
A new look into Web page ranking systems
Zin T., Tin P., Hama H., Toriu T.
Advances in Intelligent Systems and Computing 329 343 - 351 2015
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Advances in Intelligent Systems and Computing
© Springer International Publishing Switzerland 2015. This paper proposes a new way of looking into Web page ranking systems by using some concepts of queuing theory in operations research and stochastic water storage theory in hydrology. Since both theories queuing and stochastic water storage are rich in technology as well as application aspects, the new look in this paper may lead to new directions in Web page ranking systems and related research areas. In doing so, first this paper draws some analogies between a Web page ranking system and theory of queues. Then it shows how a Web page ranking system can be tackled to reduce current obstacles by using queuing theory techniques. In the second, a Web page ranking system is modeled as a framework of stochastic water storage theory to derive a list of Web page rankings. Third and finally, the outcome results of rankings obtained by using the proposed two theories queuing theory and stochastic water storage are compared and analyzed analytically as well as experimentally. The experimental results show the proposed new look is promising for establishing a new research area which can improve the current situations and difficulties occurred in search engines and their ranking systems in particular and some problems in World Wide Web as a whole.
-
A matrix-geometric method for web page ranking systems
Zin T., Tin P., Hama H., Toriu T.
Journal of Information Hiding and Multimedia Signal Processing 6 ( 4 ) 639 - 647 2015
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Journal of Information Hiding and Multimedia Signal Processing
© 2015 ISSN. An algorithmic approach which is known as Matrix-Geometric method is an efficient and popular in solving various types of stochastic models especially for queueing and water storage models. This paper proposes a modified Matrix-Geometric method for calculating random web page ranks in World Wide Web (WWW) by considering web page ranking system as a stochastic model in random environments. Since the Matrix- Geometric approach is very powerful in variety of stochastic models, it may develop new promising directions in Web Page Ranking Systems and related research areas. In order to do so as first step, a web page ranking system is modeled in the frameworks of queueing models. Then an attempt is made to reduce the obstacles occurred in the problems of web page ranking systems. Similarly, in the second step, a Web page ranking system is modeled as a framework of stochastic water storage theory to derive a list of Web page rankings by using Matrix-Geometric method. Some comparison results are presented to confirm the efficiency of the proposed methods. The experimental results shows the pro- posed approach is promising for establishing a new research area which can improve the current situations and difficulties occurred in search engines and their ranking systems in particular and some problems in WWW as a whole.
-
Zin T., Lin J., Pan J., Tin P., Yokota M.
Advances in Intelligent Systems and Computing 388 2015
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Advances in Intelligent Systems and Computing
-
A cluster based ranking framework for multi-typed information networks
Tin P., Toriu T., Zin T., Hama H.
Proceedings - 2014 10th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2014 415 - 418 2014.12
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Proceedings - 2014 10th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2014
© 2014 IEEE. A multi-typed information network is an information network which contains multiple types of objects having actions and interactions between each other. Although many studies on single typed information network haven been found in the literature, only a little has been known concerning with multi-typed information networks. On the other hand, multiple type information networks are ubiquitous and forming an important component of modern information infrastructure. Thus, in this paper we propose a new method to give a better understanding of information networks and their properties. Specifically we propose a new cluster based ranking system for multi-typed information networks. In this aspect, ranking evaluates objects of information networks based on some mathematical ranking function which illustrates the characteristic of objects with which any two objects of the same type can be compared by qualitatively. Moreover, clustering group objects is based on a certain measure such that similar objects are in the same cluster whereas dissimilar objects are in different clusters. Then the ranking and clustering processes are integrated to extract insight overall views of information networks, so that the integrated method can be widely applied in different information network settings. Our experiments using DBLP datasets can generate good informative clusters producing reliable ranking system.
-
A new background subtraction method using bivariate poisson process Reviewed
Thi Thi Zin, Pyke Tin, T. Toriu, H. Hama
Proceedings - 2014 10th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2014 419 - 422 2014.12
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Proceedings - 2014 10th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2014
Background subtraction is one of important fundamental steps in many image processing applications such as object recognition, detection, tracking, human behavior analysis in video surveillance systems, etc. So the background subtraction method must be efficiency, that is saving time and space and have a good performance. In order to achieve this aim, a new background subtraction method is proposed by using a bivariate Poisson process which takes both serial and spatial correlations of image pixels. The proposed method can deal with complex background scenarios including slowly moving foreground objects, illumination changes and etc. Numerous experiments on various types of video sequences show that the method is robust to compare with several existing methods, can achieve very promising performance.
-
A new incremental principal component analysis with a forgetting factor for background estimation Reviewed
T. Toriu, Thi Thi Zin, H. Hama
SPIE Proceedings: Electro-Optical and Infrared Systems: Technology and Applications XI 9249 - 7 2014.10
Language：English Publishing type：Research paper (international conference proceedings)
DOI： 10.1117/12.2067001
-
Image Smoothing Using a Metric Tensor for an Affine Invariant Scale Space Reviewed
T. Toriu, Thi Thi Zin, H. Hama
Proc. of 4th International Conference on Image Processing Theory, Tools and Applications (IPTA 2014) 1 - 6 2014.10
Language：English Publishing type：Research paper (international conference proceedings)
-
A Human Behavior Analyzer Framework for Consumer Product Search Engines Reviewed
Thi Thi Zin, Pyke Tin, T. Toriu, H. Hama
Proc. of the 3rd IEEE Global Conference on Consumer Electronics (GCCE 2014) 138 - 139 2014.10
Language：English Publishing type：Research paper (international conference proceedings)
-
An Embedded Knowledge Extraction Technology for Consumer Video Surveillance Reviewed
Thi Thi Zin, Pyke Tin, T. Toriu, H. Hama
Proc. of the 3rd IEEE Global Conference on Consumer Electronics (GCCE 2014) 510 - 511 2014.10
Language：English Publishing type：Research paper (international conference proceedings)
-
A Cluster Based Ranking Framework for Multi-Typed Information Networks Reviewed
Pyke Tin, Thi Thi Zin, T. Toriu, H. Hama
Proc. of 10th International Conference on Intelligent Information Hiding and Multimedia Signal Processing (IIH-MSP 2014) 415 - 418 2014.8
Language：English Publishing type：Research paper (international conference proceedings)
-
A New Background Subtraction Method using Bivariate Poisson Process Reviewed
Thi Thi Zin, Pyke Tin, T. Toriu, H. Hama
Proc. of 10th International Conference on Intelligent Information Hiding and Multimedia Signal Processing (IIH-MSP 2014) 419 - 422 2014.8
Language：English Publishing type：Research paper (international conference proceedings)
-
A Novel Method for Contact-Free Measurement of Dolphin's Body Size Using a 3D Bezier Curve Reviewed
H.Hama, T. Morisaka, Thi Thi Zin, Y. Hashimoto, R. Matsui
ICIC Express Letters (Part B: Applications): An International Journal of Research and Surveys 5 ( 2 ) 583 - 587 2014.4
Language：English Publishing type：Research paper (scientific journal)
-
A Data-Driven Key Information Search System in Big Data Analytics Reviewed
Pyke Tin, Thi Thi Zin, H. Hama, T. Toriu
ICIC Express Letters (Part B: Applications): An International Journal of Research and Surveys 5 ( 2 ) 365 - 370 2014.4
Language：English Publishing type：Research paper (scientific journal)
-
A Correlated Random Walk First Passage Time Distance for Data Clustering with Medical Applications Reviewed
Thi Thi Zin, Pyke Tin, T. Toriu, H. Hama
ICIC Express Letters (Part B: Applications): An International Journal of Research and Surveys 5 ( 2 ) 577 - 582 2014.4
Language：English Publishing type：Research paper (scientific journal)
-
A correlated random walk first passage time distance for data clustering with medical applications
Zin T., Tin P., Toriu T., Hama H.
ICIC Express Letters, Part B: Applications 5 ( 2 ) 577 - 582 2014.2
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
A distance measure for a given dataset plays an important role in clustering processes and computing communities in most real world networks including medical data analysis. In this paper, we propose a correlated random walk first passage time measure of distances between data points or nodes in order to produce an effective clustering algorithm with respect to medical applications. In order to do so, we first model the correlated random walk on the graph by using a mixture of gamma distributions as weight function. We then compute the expected first passage time for the random walk to travel between two nodes and return which are used as distance measures for clustering data points. The performance evaluation of the proposed method is carried out by using some medical benchmark datasets with comparison to other methods. © 2014 ISSN 2185-2766.
-
A novel method for contact-free measurement of dolphin's body size using a 3D Béezier curve Reviewed
H. Hama, T. Morisaka, Thi Thi Zin, Y. Hashimoto, R. Matsui
ICIC Express Letters, Part B: Applications 5 ( 2 ) 583 - 587 2014.2
Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
This paper proposes a novel method for contact-free measurement of an object's length with non-rigid deformation using a 3D Bézier curve. This technique is developed to measure body size of the free-ranging wild dolphins without touching them. Body size is the most basic and important information to measure the health condition of an animal population. We adopt here a new model for dolphin's body movement to develop a theoretical algorithm to realize it. This technique will contribute not only to know dolphin's body size but to conserve wild dolphins. The result of preliminary experiment using our contact-free measurement of 3D curve length shows 2% or less error, which is the worthy scenario for field study, even if we must take into account limited underwater visibility, a change of refractive index and so on.
-
A data-driven key information search system in big data analytics
Tin P., Zin T., Hama H., Toriu T.
ICIC Express Letters, Part B: Applications 5 ( 2 ) 365 - 370 2014.2
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
Nowadays, the emergence of data-driven information trends has exploded the volume, velocity and variety referring to as Big Data. Analyzing and understanding the Big Data for key information trends and unusually high or low activity is an important problem in the Big Data analytics having many applications. In this paper, we propose a framework for a data-driven key information search system in Big Data analytics. Since many of the Big Data problems are dependent on the unstructured data analysis, the proposed framework is based on an arbitrarily related data records and metric spaces to study such problems. In doing so, a set theoretic data structure is employed for expressing non-intrinsically related massive large data sets and transforming into a stochastic model to compute a new stream of right information for right targets at right time. By processing a steady state distribution of the model based on real-time data, the framework enables to make time-sensitive decisions, monitor emerging trends, and jump on new insight information. As an illustration, the unstructured data targeted in this work to organize, is the public consuming patterns available in social network platforms and in a kind of various types of sensor readings, followed by time series data analysis to obtain the necessary key information over Big Data. © 2014 ISSN 2185-2766.
-
An embedded knowledge extraction technology for consumer video surveillance Reviewed
Thi Thi Zin, Pyke Tin, T. Toriu, H. Hama
2014 IEEE 3rd Global Conference on Consumer Electronics, GCCE 2014 510 - 511 2014.2
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2014 IEEE 3rd Global Conference on Consumer Electronics, GCCE 2014
New advances in embedded computing technology have opened up the potential for new era of consumer surveillance systems. This paper will explore and propose a new embedded modeling technique for the configuration of consumer video surveillance systems that can identify events of interest, especially on abandoned and stolen objects in indoor and outdoor environments. The proposed embedded system will focus on high level behavior understanding for object detection, tracking and classification. The experimental results illustrate the ability of the system to create complex spatiotemporal relations and to recognize the behavior of one or multiple objects in various video scenes.
-
An integrated framework for detecting suspicious behaviors in video surveillance
Zin T., Tin P., Hama H., Toriu T.
Proceedings of SPIE - The International Society for Optical Engineering 9026 2014
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Proceedings of SPIE - The International Society for Optical Engineering
In this paper, we propose an integrated framework for detecting suspicious behaviors in video surveillance systems which are established in public places such as railway stations, airports, shopping malls and etc. Especially, people loitering in suspicion, unattended objects left behind and exchanging suspicious objects between persons are common security concerns in airports and other transit scenarios. These involve understanding scene/event, analyzing human movements, recognizing controllable objects, and observing the effect of the human movement on those objects. In the proposed framework, multiple background modeling technique, high level motion feature extraction method and embedded Markov chain models are integrated for detecting suspicious behaviors in real time video surveillance systems. Specifically, the proposed framework employs probability based multiple backgrounds modeling technique to detect moving objects. Then the velocity and distance measures are computed as the high level motion features of the interests. By using an integration of the computed features and the first passage time probabilities of the embedded Markov chain, the suspicious behaviors in video surveillance are analyzed for detecting loitering persons, objects left behind and human interactions such as fighting. The proposed framework has been tested by using standard public datasets and our own video surveillance scenarios. © 2014 SPIE.
DOI： 10.1117/12.2041232
-
New information search system in social networks application to disaster event analysis
Zin T., Tin P., Toriu T., Hama H.
IAENG Transactions on Engineering Sciences - Special Issue of the International MultiConference of Engineers and Computer Scientists, IMECS 2013 and World Congress on Engineering, WCE 2013 411 - 419 2014
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：IAENG Transactions on Engineering Sciences - Special Issue of the International MultiConference of Engineers and Computer Scientists, IMECS 2013 and World Congress on Engineering, WCE 2013
This chapter is concerned with the development of a new information search system in social network platforms for analyzing events in case of natural and man-made disasters. In todays, the world has witnessed the role of many social networks such as Facebook, Flicker, Twitter, and YouTube for extracting key and crucial information when worldwide scale major events had occurred. A social network can also be considered as information link creator between users and available resources. This role has been boosted in tremendous amount during and after the recent disasters around the world; for example, the 2010 Philippine typhoon, the 2010 Haiti earthquake, the 2011 Brazil flood, and the 2011 Japan earthquake and tsunami and Boston Marathon Explosion 2013. In such emergency situations, extracting and analyzing key information from the congested information resources posted or hosted in variety of forms and norms through almost all social network platforms are of major concerns in assessing the situation and in decision making. Therefore, some mathematical modeling techniques of branching processes and Markov chain theory to investigate how new knowledge and new information about the disasters spreads on the social networks and how to extract trust and reliable key and dominant information are discussed here. Specifically, a set of text messages and visual information is transformed into a Markov chain to produce stationary and time dependent distributions. Then, the abnormalities and suspicious patterns occurring in the distributions are analyzed for detecting and extracting key information. Finally, some simulation results are given in the case of Boston Marathon Explosion occurred in April 15, 2013 by using three social networks Flickr, YouTube and Twitter information. © 2014 Taylor & Francis Group, London.
-
A Big Data application framework for consumer behavior analysis Reviewed
Thi Thi Zin, Pyke Tin, T. Toriu, H. Hama
2013 IEEE 2nd Global Conference on Consumer Electronics, GCCE 2013 245 - 246 2013.12
Language：English Publishing type：Research paper (international conference proceedings) Publisher：2013 IEEE 2nd Global Conference on Consumer Electronics, GCCE 2013
More than ever before, the amount of data about consumers, suppliers and products has been exploding in today consumer world referred as 'Big Data'. In addition, more data is available to the consumer world from multiple sources including social network platforms. In order to deal with such amount of data, a new emerging technology 'Big Data Analytics' is explored and employed for analyzing consumer behaviors and searching their information needs. Specifically, this paper proposes a Big Data application framework for analyzing consumer behaviors by using topological data structure, co-occurrence methodology and Markov chain theory. First, the consumer related data is translated into a topological data structure. Second, using topological relationships, a co-occurrence matrix is formed to deduce Markov chain model for consumer behavior analysis. Finally, some simulation results are shown to confirm the effectiveness of the proposed framework. © 2013 IEEE.
-
Archery sight-system by magnetic sensors for visually impaired persons
Sasayama T., Oka I., Ata S., Zin T., Watanabe H., Sasano H.
International Conference on Advanced Technologies for Communications 559 - 562 2013.12
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：International Conference on Advanced Technologies for Communications
The archery sight-system is developed for visually impaired persons. The system is based on the magnetic directional sensors for earth magnetism. Both azimuth and elevation angles to the target are measured by the sensor equipped on the archery bow. The sensed information is transmitted to the receiver, which is located near the archer. The sound interface in the receiver transforms the sensed information to the hearing information, and the archer gets this hearing information by the wireless headphones. The sensor system is light-weight of 34g, and operates for about 2 hours by two SR44 button batteries. Shooting experiments by archers with eye-mask show that the system is useful for the visually impaired persons to play archery. © 2013 IEEE.
-
A stochastic model for measuring popularity and reliability in social network systems
Zin T., Tin P., Toriu T., Hama H.
Proceedings - 2013 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2013 462 - 467 2013.12
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Proceedings - 2013 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2013
Popularity and reliability information are crucial ingredients of today social networking systems such as Facebook, LinkedIn, YouTube, Twitter, and so on. In this paper, we propose a stochastic model for measuring and ranking popularity and reliability information in social networks. Specifically, by using the relationships between co-occurring users, we model a Markov chain for reliability measures and a queuing model for popularity measures based on time, space and scenarios. We then form a convex combination or fusion of the two measures to compute the Integrated Global Rank for social networks systems. Finally, we present some illustrative simulation results by using the social networks data collected from Twitter and YouTube. Experimental study indicates the effectiveness of proposed ranking algorithm in terms of better search results. © 2013 IEEE.
DOI： 10.1109/SMC.2013.84
-
Similar image retrieval system reflecting user intention using shape descriptors
Iwanaga T., Zin T., Hama H., Toriu T., Tin P.
ICIC Express Letters 7 ( 4 ) 1425 - 1430 2013.2
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters
Most existing image search systems cannot always provide satisfactory information to the users because relevant images differ according to user or requirement. It is a challenging task to establish an image search system which effectively and properly takes users demands into account. Thus in this paper we propose a new technique for searching images in large databases to meet the user's demands. The proposed system reflects user demand by user selection from choices showed by system. In our system, we used two shape descriptors: the region-based and contour-based. Moreover, we confirmed the effectiveness of the proposed system by comparing with the ordinary system. Experimental results show that our approach is very powerful and this also leads to satisfy the user satisfaction. © 2013 ICIC International.
-
A two dimensional correlated random walk model for visual tracking
Zin T., Tin P., Toriu T., Hama H.
ICIC Express Letters 7 ( 5 ) 1501 - 1506 2013.2
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters
This paper proposes a novel approach to visual tracking by using a sequential two dimensional correlated random walk model. In this approach we use Markov chain sampling techniques to estimate the most likely parameters including the spatiotemporal characteristics of object motions in the setting of occlusions and clutters. Based on the estimated parameters, a discriminant probability matrix is established to extract the target foreground object. The probability matrix is updated by an incremental maximal principle to handle the appearance variations of the target and background. The fundamental equation which describes a two dimensional random walk with non-uniform jump is derived by using the updated probability matrix. Then, it is applied to a visual analysis of human motion tracking problems. The efficiency of the proposed method is confirmed by our experimental results obtained in complex scenario. © 2013 ISSN 1881-803X.
-
A stochastic petri net framework for human behavior analysis in surveillance video Reviewed
Pyke Tin, Thi Thi Zin, T. Toriu, H. Hama
ICIC Express Letters 7 ( 5 ) 1675 - 1680 2013.2
Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters
Analyzing human behaviors and detecting unusual events is an essential task in surveillance applications for prevention of security threats. In order to carry out this task, a proper way of representing an event and an efficient method of recognizing such events are required. In this paper, we propose a stochastic Petri Net framework for both representation and recognition of unusual events involving suspicious people and suspicious objects. We first use a formal low level image processing method for background subtraction and then automatically map the extracted foregrounds into stochastic Petri Net framework for event representations. In the next we generate reachability graphs to describe human behavior models concerning with specific event nets. Recognition process is then carried out by moving tokens through the Petri Nets. In particular we investigate some unusual human-object-human interactions such as theft detection, exchange of objects and two persons fighting. The experimental results show the effectiveness of our approach to human behavior analysis problems attaining high accuracy rates. © 2013 ISSN 1881-803X.
-
A series of action recognition using HMM and probability representation based on binned features set
Sugimoto M., Zin T., Toriu T., Tin P.
ICIC Express Letters 7 ( 4 ) 1419 - 1424 2013.2
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters
We present a more practical approach for human action recognition in video image sequences. The methodology is the combination of Hidden Markov Model (HMM) and probability representation based on binned features set, including silhouette and feature extraction. This enables to trace a person's action consecutively. For this aim, we define seven types of fundamental actions such as walking and running. The proposed system consists of two major steps: the training of HMM parameters and the testing video sequences for a series of actions. The experimental results show that the system can recognize the seven actions with high accuracy rate. Simultaneously, they also show HMM produces good performance to serial human action recognition for given the features set and training sequences. © 2013 ICIC International.
-
Image processing approach to automatic scoring system for archery targets
Zin T., Oka I., Sasayama T., Ata S., Watanabe H., Sasano H.
Proceedings - 2013 9th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2013 259 - 262 2013
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Proceedings - 2013 9th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2013
This paper proposes an image processing approach to automatic scoring system for archery targets. Specifically, the proposed system first performs Morphological operations on the target image to obtain thick boundaries of the arrow hits and then the image segmentation process is applied to segment the target area by using color and shape features. In doing so, a dynamic thresholding is applied to handle the possible effects of illumination variations and the bull's eyes or the target concentric circles are segmented by using geometrical properties of circles and ellipses to calculate hit scores. Thus the system can have a capability of scoring inside and outside the concentric circles separately. The proposed method is tested by using our own video sequences taken in the self-conducted experiments performed by four archers on two different days shooting 6 arrows. The test results show that the proposed scoring system is promising with an accuracy rate of 100%. © 2013 IEEE.
-
An integrated framework for disaster event analysis in big data environments
Tin P., Zin T., Toriu T., Hama H.
Proceedings - 2013 9th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2013 255 - 258 2013
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Proceedings - 2013 9th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2013
Today world has witnessed the catastrophic consequences of natural and man-made disasters are demanding the urgent need for more research to advance fundamental knowledge and innovation for disaster prevention, mitigation and management. At the same time, the world is in the age of the Big Data revolution which holds the potential to mitigate the effects of disaster events by enabling access to critical real time information. Thus, in this paper an integrated framework for analyzing disaster events by using the Big Data analytics is proposed. The proposed framework shall address three key components to perform data organization, data integration and analysis, information presentation to users by utilizing Big Data with respect to disaster events. In doing so, the paper shall create a disaster domain-specific search engine using co-occurring theory and Markov chain concepts for preparing impacts of disaster attacks to make the society better aware of the situations. Specifically, stochastic clustering with constraints is used to automatically extract disaster events by defining the set of structural attributes. Some illustrative simulations are shown by using Big Data sources for the Great East Japan earthquake, tsunami and nuclear disaster events of 2011. © 2013 IEEE.
-
A Markov chain model for image ranking system in social networks
Zin T., Tin P., Toriu T., Hama H.
Proceedings of SPIE - The International Society for Optical Engineering 9027 2013
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Proceedings of SPIE - The International Society for Optical Engineering
In today world, different kinds of networks such as social, technological, business and etc. exist. All of the networks are similar in terms of distributions, continuously growing and expanding in large scale. Among them, many social networks such as Facebook, Twitter, Flickr and many others provides a powerful abstraction of the structure and dynamics of diverse kinds of inter personal connection and interaction. Generally, the social network contents are created and consumed by the influences of all different social navigation paths that lead to the contents. Therefore, identifying important and user relevant refined structures such as visual information or communities become major factors in modern decision making world. Moreover, the traditional method of information ranking systems cannot be successful due to their lack of taking into account the properties of navigation paths driven by social connections. In this paper, we propose a novel image ranking system in social networks by using the social data relational graphs from social media platform jointly with visual data to improve the relevance between returned images and user intentions (i.e., social relevance). Specifically, we propose a Markov chain based Social-Visual Ranking algorithm by taking social relevance into account. By using some extensive experiments, we demonstrated the significant and effectiveness of the proposed social-visual ranking method.
DOI： 10.1117/12.2042621
-
Knowledge based social network applications to disaster event analysis
Zin T., Tin P., Hama H., Toriu T.
Lecture Notes in Engineering and Computer Science 2202 279 - 284 2013
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Lecture Notes in Engineering and Computer Science
Today online social networking platforms such as Facebook, Flicker, Twitter, and YouTube often serve a breaking news role for natural disasters. The role of these social networks has significantly increased following the recent disasters around the world; the 2010 Philippine typhoon, the 2010 Haiti earthquake, the 2011 Brazil flood, and the 2011 Japan earthquake and tsunami. Moreover, these platforms are among the first ones to help communicate the news to a large mass of people since they are visited by millions of users regularly. In such emergency situations, detecting and analyzing hot spots or key events from the pool of information in the social networks are of major concerns in assessing the situation and in decision making. In this paper, a knowledge based event analysis framework for automatically analyzing key events is proposed by using various social network sources in case of disasters. In doing so, some mathematical modeling techniques of branching processes and Markov chain theory are explored and employed to investigate how news about these disasters spreads on the social networks and how to extract trust and reliable key information. Specifically the abnormal or suspicious topics and important events within various social network platforms are analyzed by using a set of selected messages and visual data. Finally some illustrative sample results are presented based on a limited datasets of YouTube and Twitter in the case of March 11, 2011 Japan Earthquake and Tsunami.
-
Visual behavior analysis tool for consumer video surveillance
Zin T., Tin P., Toriu T., Hama H.
1st IEEE Global Conference on Consumer Electronics 2012, GCCE 2012 718 - 719 2012.12
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：1st IEEE Global Conference on Consumer Electronics 2012, GCCE 2012
Consumers video surveillance systems are now being used not only for security reasons but also for better understanding consumer behaviors. In this paper, we propose a new visual behavior analysis tool for consumer video surveillance systems. This tool can be embedded in consumer videos to automatically detect and analyze unusual events. The proposed tool is developed by using a special type of Gamma Markov chain for background modeling and Petri Nets for object classification. We present some experimental results to show the effectiveness of the proposed system which will be leading to new visual behavior analysis tools for the consumers. © 2012 IEEE.
-
Conceptual vision keys for consumer product images
Zin T., Tin P., Toriu T., Hama H.
1st IEEE Global Conference on Consumer Electronics 2012, GCCE 2012 435 - 436 2012.12
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：1st IEEE Global Conference on Consumer Electronics 2012, GCCE 2012
In the consumer world, the ever growing image repositories in online shopping, consumer products images, consumer photos and video collections have resulted great demand of a system which can accurately retrieve similar images from image database. For this purpose, we propose a new concept of vision key for retrieving consumer product images. In our system, rather than considering an image as a whole, we consider it as a set of regions or sub-images with completely different semantic meanings. By using the properties of equivalence classes in the Markov chain, we first perform image segmentation and initial pixel grouping process. We then establish vision keys by using a Markov stationary feature. Finally, in the retrieval phase, users can interactively search candidate images which contain vision keys. In order to confirm the efficiency of our proposed method, we present the experimental results achieving on higher accuracy rates. © 2012 IEEE.
-
A series of stochastic models for human behavior analysis
Zin T., Tin P., Toriu T., Hama H.
Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics 3251 - 3256 2012.12
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics
In this paper, we propose a new approach to analyze human behaviors by using a series of stochastic models composed of a Bivariate Gamma Markov model, a two dimensional correlated Random Walk model and a finite state Markov Chain model. Specifically, the proposed method contains three modules namely: (i) image analysis module, (ii) probability analysis module and (iii) event analysis module. We model each module by a special type of stochastic processes forming a series of stochastic models for the complete behavior analysis system. This approach is more effective in utilizing modular stochastic models to describe complex behavior patterns. By assembling these modular models in a series we can design a robust model for the analysis of human behaviors. The feasibility and effectiveness of the proposed method are tested on two different datasets: a self-collected dataset and PETS 2006 dataset. © 2012 IEEE.
-
A novel probabilistic video analysis for stationary object detection in video surveillance systems
Zin T., Tin P., Toriu T., Hama H.
IAENG International Journal of Computer Science 39 ( 3 ) 295 - 306 2012.9
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：IAENG International Journal of Computer Science
In this paper, we propose a novel probabilistic approach for detecting and analyzing stationary objects driven visual events in video surveillance systems. This approach is based on a newly developed background modeling technique and an adaptive statistical sequential analysis method. For background modeling part, we use the concepts of periodic Markov chain theory producing a new background subtraction method in computer vision systems. We then develop an object classification algorithm which can not only classify the objects as stationary or dynamic but also eliminate the unnecessary examination tasks of the entire background regions. Finally, this paper introduces a sequential analysis model based on exponent running average measure to analyze object involved events such as whether it is either abandoned or very still person. In order to confirm our proposed method we present some experimental results tested on our own video sequences taken in international airports and some public areas in a big city. We have found that the results are very promising in terms of robustness and effectiveness.
-
Motion-compensated inter-frame subtraction based on self-organized internal representation Reviewed
T. Toriu, Thi Thi Zin, H. Hama
ICIC Express Letters 6 ( 4 ) 905 - 910 2012.4
Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters
In a previous paper, we introduced a time evolution operator and based on it proposed an unsupervised learning algorithm for self-organizing an internal representation of ego-motion. In this paper, we propose a method to predict the image at the next instance using the time evolution operator generated on the basis of the internal representation of ego-motion. In addition, we propose a method of motion-compensated inter-frame subtraction. By subtracting the predicted image at the next instance from the true image, we can obtain an image that has high intensity in the region of the moving object. This method is effective even if the camera itself is in motion. We show the results of the experiments conducted using a randomly synthesized image and the real image. ICIC International © 2012.
-
An innovative background model based on multiple queuing framework Reviewed
Thi Thi Zin,Pyke Tin, T. Toriu, H. Hama
ICIC Express Letters 6 ( 4 ) 1039 - 1044 2012.4
Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters
Many computer vision applications, especially video surveillance systems, highly depend on background model and foreground segmentation. In this paper, we propose an innovative background subtraction method based on a multiple queuing framework. Under this framework, both the static and dynamic background pixels are investigated by using a novel hypothesis method to establish an active real time background modeling in presence of moving foreground objects in the complex scene and adaptation of background model to gradual and sudden "once-off" background changes. Experiments were conducted by using public datasets PETS2006 and our own video sequences taken at an international airport and a university campus.
-
A general framework for knowledge based human behavior understanding Reviewed
Pyke Tin, Thi Thi Zin, T. Toriu, H. Hama
ICIC Express Letters 6 ( 4 ) 899 - 904 2012.4
Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters
Human behavior understanding systems today are mainly knowledge based. And the diffusion of human actions and interactions knowledge is imperative for proper and prompt decision making processes in human behavior related crime prevention, care support systems, intelligent surveillance systems, etc. In this paper, we propose a general framework for knowledge based human behavior understanding system comprising three components: (i) knowledge acquisition, (ii) knowledge representation and modeling, and (in) knowledge base and use of knowledge. Specifically, human behavior is modeled as a Stochastic sequence of low level actions. The results of the Low Level Activities (LLA) analysis will subsequently be fed into a Knowledge Base System (KBS) that is used as High Level Activities (HLA) model. As an application, we will focus on the detection of theft and robbery related events. Unlike the traditional approach to just detecting stationary and moving objects in monitored scenes, our KBS approach detects the events based on accumulated knowledge about human and non-human objects from continuous object classification. ICIC International © 2012.
-
A probability-based model for detecting abandoned objects in video surveillance systems
Zin T.
Lecture Notes in Engineering and Computer Science 2198 1246 - 1251 2012
Language：English Publishing type：Research paper (scientific journal) Publisher：Lecture Notes in Engineering and Computer Science
© 2012 Newswood Limited. All rights reserved. Detection of suspicious packages or abandoned objects is one of the most important tasks in video surveillance systems. Some recent terrorist attacks involving explosive packages left behind in many contexts such as airports, rail stations and etc. illustrate the importance of this problem. In this paper, we propose a probability-based model for robustly and efficiently detecting abandoned objects in complex environments. Specifically, we develop a new probability-based background subtraction algorithm based on combination of multiple background models for motion detection. In addition, several improvements are implemented to the background subtraction method for shadow removal and quick lighting change adaptation. We then analyze the extracted objects to classify as static or dynamic objects. After the analysis, we employ the statistical running average of the static foreground masks for event type decision making either abandoned or very still person. Finally, the robustness and efficiency of the method are tested on our video sequences and PETS2006 datasets.
-
Simultaneous visual ranking and clustering using weighted multiple features
Geng T., Zin T., Tin P., Toriu T., Hama H.
ICIC Express Letters 5 ( 10 ) 3773 - 3778 2011.10
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters
In this paper, we present a new approach to visual ranking and clustering system by jointly exploring the multiple visual features information using adaptive visual similarity. We investigate how to effectively incorporate both intra-image and inter-image spatial structure information into Markov Stationary Features derived from the normalized co-occurrence matrix. The convex quadratic programming algorithm is developed to learn the weights for fusing the rankings from multiple features using color and texture. As a result, the multiple visual features based re-ranking can take more reliable information from each other. Experimental results on a real-world datasets collected from various image search engines show that our method outperforms several existing approaches which do not or weakly consider multi-feature interactions. © 2011 ISSN 1881-803X.
-
Zin T., Tin P., Hama H., Nakajima S., Toriu T.
ICIC Express Letters 5 ( 10 ) 3767 - 3772 2011.10
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters
In this paper, we present a robust and effective multiple layers stochastic background modeling and novel stationary object detection method, which comprise of probability based filtering operations to detect stationary objects in a monitoring scene. Conventionally, a statistical background model is extracted by using a training sequence without foreground objects. Our method does not require starting with a period of empty scenes to facilitate the original background. In our proposed algorithm, three layers background model is constructed by using periodic Markov chain concepts. We then apply background subtractions to the current frame for objects detection and classification. Extensive experimental work has been done, results of which show that the present approach provides a better solution compared with the conventional approach, including the problem of re-active objects in real world complex environments. © 2011 ISSN 1881-803X.
-
Challenges and promises in human behavior understanding research
Tin P., Zin T., Hama H., Toriu T.
ICIC Express Letters 5 ( 10 ) 3761 - 3766 2011.10
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters
Visual surveillance for human behavior understanding is an active research topic in image processing. Behavior analysis in dynamic scenes is a complex task, especially when it concerns human populated environments. Study to emulate the astonishing performances of such a perfect system as the natural and computer vision system represents, without any doubt, a real challenge. It has a wide spectrum of promising applications, including access control in special areas, human identification at a distance, crowd flux statistics and congestion analysis, detection of anomalous behaviors, and interactive surveillance using multiple cameras, etc. In this paper, the current state-of-the-art image processing methods for automatic behavior recognition techniques are described, with a focus on the surveillance of human activities in the context of transit applications. The main purpose of this paper is to provide researchers in the field with a summary of progress achieved to date and to help identify areas where further research is needed. This paper presents a comprehensive study of the research on relevant human behavior understanding methods for public safety and security surveillance. A classification table of research papers on relevant behavior analysis is presented, including behaviors, datasets, implementation details, and results. © 2011 ISSN 1881-803X.
-
The method of area segmentation of cancer using multispectral image
Takahashi H., Yamada K., Tareda M., Yoshida S., Zin T., Aida T.
ICIC Express Letters 5 ( 10 ) 3859 - 3864 2011.10
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters
Transnasal endoscopy has recently become a widely accepted screening method in practical application areas of medical endoscopy. Pain tolerance and safety seem to be the greatest advantage of transnasal endoscopy when compared with those of a conventional peroral endoscopy. Although there are some disadvantages such as poor endoscopic images and less illumination caused by downsizing of scopes, as a whole, slim endoscopy is a safe and less invasive tool for screening purposes. On the other hand, flexible spectral imaging color enhancement (FICE) is one of the diagnostic methods using specific light spectra based on spectral image processing technology. FICE provides comparison of spectral images of diseased and surrounding normal areas for enhancement of the contrast by combining wavelengths with greater differences in signals. It is thus in this paper, that we propose a novel method of segmenting tumor region for medical endoscope surgery and investigation by using the narrowband images on several wavelengths. © 2011 ISSN 1881-803X.
-
Small sample learning for motion estimation based on self-organized internal representation
Toriu T., Zin T., Hama H.
ICIC Express Letters 5 ( 10 ) 3921 - 3926 2011.10
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters
In the previous paper, we proposed an unsupervised learning algorithm for self-organizing an internal representation of ego-motion. We showed that motion param-eters could be topographically mapped onto a robot's internal parameter space sponta-neously. In this paper, first, we provide the theoretical rationale why the previous method can self-organize the internal representation of ego-motion. Then, on the basis of this theoretical foundation, we propose a novel learning method to estimate real motion parameters such as translation and rotation parameters. Only small samples of input and output data are needed to complete this learning. We show that this method works well by experiments using randomly synthesized image. © 2011 ISSN 1881-803X.
-
Pedestrian detection based on hybrid features using near infrared images
Zin T., Tin P., Hama H.
International Journal of Innovative Computing, Information and Control 7 ( 8 ) 5015 - 5025 2011.8
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：International Journal of Innovative Computing, Information and Control
This paper explores a hybrid-based method to fuse multi-slit features and Histograms of Oriented Gradients (HOG) features for pedestrian detection from Near Infrared (NIR) images. The fused feature set utilizes both the multi-slit method's capability of accurately capturing the local spatial layout of body parts (head, torso and legs) in individual frames and the HOG's capability in region information relevant to higher frequency components. The hybrid feature vector describing various types of poses is then constructed and used for detecting the pedestrians. The part based pattern matching analysis indicates that the fused features have much higher feature space separation than the pure features. Experiments with a database of NIR images show that the proposed method achieves a substantial improvement in tackling some difficult cases such as side view, back view which the conventional HOG method cannot handle. Detection and recognition performance is less computationally expensive than existing approaches. © 2011 ICIC International.
-
Background modeling using special type of Markov Chain
Zin T., Tin P., Toriu T., Hama H.
IEICE Electronics Express 8 ( 13 ) 1082 - 1088 2011.8
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：IEICE Electronics Express
Background modeling is important in video surveillance for extracting foreground regions from a complex environment. In this paper, we present a novel background modeling technique based on a special type of Markov Chain. The method is a substantial extension to the existing background subtraction techniques. First, a background pixel is statistically modeled by a linear regressive Gamma Markov distribution. Then, these statistical estimates are used as important parameters in background update schemes. The experimental results show that the proposed model is less sensitive to movements of the texture background and more robust for real time segmenting the foreground object accurately. © IEICE 2011.
DOI： 10.1587/elex.8.1082
-
A novel concept of morphology pivotal elements for object image retrieval
Hama H., Zin T., Tin P.
International Journal of Innovative Computing, Information and Control 7 ( 7 A ) 3891 - 3901 2011.7
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：International Journal of Innovative Computing, Information and Control
In this paper, we introduce a novel and simple Pivotal Element (PE) concept in mathematical morphological operations for object image retrieval schemes based on combinations of empirical and statistical analyses. Mathematical morphology is very attractive for this purpose because it efficiently deals with geometrical features like size, shape, contrast or connectivity that can be considered as image retrieval oriented features. With an optimized structure, morphological dilation is more effective to detect object spot target in image sequences. Based on the real convex figure, morphological operation with circular structure is designed in this paper. The PE is introduced to optimize the noisy background elements. The empirical threshold is decided approximately based on the statistical characters. In this aspect, two approaches for solving morphological applications to image data distributed on the unit circle are presented. In the first approach, a framework for analyzing images, called pivotal role, has been developed based on a set of concentric circles with adjustable radii, with exactly one circle centered at each pivotal image pixel. The second approach is based on Markov decision processes which operate only on grouped data. The retrieval quality is improved by dynamically changing the combinatorial coefficients that are used in equations of optimality principles. by using it as a priori knowledge of the morphology operation, it does favor to improve the algorithm's accuracy and adaptability. The experiment shows that the new concept of PE has made the morphological operations to achieve a higher retrieval efficiency and accuracy. © 2011.
-
Fast multi-slit composition method using attitude measurement unit for pedestrian detection
Hashimoto Y., Zin T., Toriu T., Hama H.
ICIC Express Letters, Part B: Applications 2 ( 3 ) 541 - 546 2011.6
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
During the next decade, on-board pedestrian detection systems will play a key role in the challenge of increasing traffic safety. The main goal of these systems, to detect pedestrians in urban scenarios, involves overcoming difficulties like processing outdoor scenes from a mobile platform and searching for aspect-changing objects in cluttered environments. Such systems thus combine various techniques in state-of-the-art computer vision. In this paper, we present a novel method that uses an Attitude Measurement Unit (AMU) in conjunction with a camera for tracking and which performs well in dynamic and cluttered outdoor environments as long as the target occlusions and losses are temporary. The main purpose is to demonstrate the advantages of using an AMU as a means of on-road pedestrian detection for automobiles over more conventional means. Specifically, we present a two module systems based on both 2D and 3D sensor cues. The first module uses fipitch and roll angles for calculating the vanishing line on the image to select a coherent set of regions of interest (ROIs) to be further analyzed. The second module develops a modified multi-slit method to classify the incoming ROIs into pedestrian and non-pedestrian in order to refine the final results. Our results indicate the integration of the proposed techniques gives rise to a promising system. © 2011 ICIC International.
-
Unsupervised learning algorithm for self-organizing internal representation of ego-motion Reviewed
T. Toriu, Thi Thi Zin, H. Hama
ICIC Express Letters, Part B: Applications 2 ( 3 ) 559 - 564 2011.6
Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
In this paper, we propose an unsupervised learning algorithm for self-organi-zing an internal representation of ego-motion. In this method, the system self-organizes the internal space for representing ego-motion without employing a supervisor. We es-tablish in our experiments that objective ego-motion parameters can be topographically mapped onto a robot's internal parameter space spontaneously using this learning. Once the learning is complete, the system can recall the representation of ego-motion from the pair of input sensation and its time derivative. One important aspect of this method is that the system does not use any knowledge of geometrical nature during image gen-eration; therefore, it is not affected by any image distortion such as that induced in omnidirectional or fish eye cameras.
-
Manipulation-Killer web information ranking System
Tin P., Zin T., Toriu T., Hama H.
ICIC Express Letters, Part B: Applications 2 ( 3 ) 523 - 528 2011.6
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
Link analysis has been widely used to evaluate the importance of web pages. Page-Rank, the most famous link analysis algorithm, offers an effective way to rank the pages. However, the algorithm ignores four facts. First, nowadays the way that users retrieve information is different from the previous way when web search engine was not extensively used. Second, inter-site links and intra-site links should not be treated equally. A link from a different site is more important for a page than that within the same site. Third, most users start their browsing from a homepage, which should be given more weight than other pages. Finally, most of existing ranking systems for linked networks can be manipulated by spammers who strategically place links. In this paper, we pro-pose a novel ranking system called manipulation-killer rank (MK-Rank) as a solution to these problems. Specifically, we show that expected passage time to reach a web page in a random walk measures essentially the same quantity as Page-Rank, whereas it does not depend on the manipulated in-links. We also show that it resists tampering by individuals or groups who strategically place manipulated links. In addition, we present an algorithm to efficiently compute passage time for all nodes in a massive graph; conventional algo-rithms do not scale adequately. Experimental results show that our MK-Rank algorithm outperforms other famous ranking algorithms, including Page-Rank and Traffic Rank, especially on sites recommendation and web spam avoidance. © 2011 ICIC International.
-
Unattended object intelligent analyzer for consumer video surveillance
Zin T., Tin P., Hama H., Toriu T.
IEEE Transactions on Consumer Electronics 57 ( 2 ) 549 - 557 2011.5
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：IEEE Transactions on Consumer Electronics
Consumer video camera surveillance with the continuous advancements of image processing technologies is emerging for consumer world of applications. Technology for detecting objects left unattended in consumer world such as shopping malls, airports, railways stations has resulted in successful commercialization, worldwide sales and the winning of international awards. However, as a consumer video application the need is now greater than ever for a surveillance system that is robustly and effectively automated. In this paper, we propose an intelligent vision based analyzer for semantic analysis of objects left unattended relation with human behaviors from a monocular surveillance video, captured by a consumer camera through cluttered environments. Our analyzer employs visual cues to robustly and efficiently detect unattended objects which are usually considered as potential security breach in public safety from terrorist explosive attacks. The proposed system consists of three processing steps: (i) object extraction, involving a new background subtraction algorithm based on combination of periodic background models with shadow removal and quick lighting change adaptation,(ii) extracted objects classification as stationary or dynamic objects, and (iii) classified objects investigation by using running average about the static foreground masks to calculate a confidence score for the decision making about event (either unattended or very still person). We show attractive experimental results, highlighting the system efficiency and classification capability by using our real-time consumer video surveillance system for public safety application in big cities. © 2011 IEEE.
-
A Markov random walk model for loitering people detection
Zin T., Tin P., Toriu T., Hama H.
Proceedings - 2010 6th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIHMSP 2010 680 - 683 2010.12
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Proceedings - 2010 6th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIHMSP 2010
Today video surveillance systems are widely used in public spaces, such as train stations or airports, to enhance security. In order to observe large and complex facilities a huge amount of cameras is required. These create a massive amount of data to be analyzed. It is therefore crucial to support human security staff with automatic surveillance applications, which will create an alert if security relevant events are detected. This way video surveillance could be used to prevent potentially dangerous situations, instead of just being used as forensic instrument, to analyze an event after it happened. In this treatise we present a surveillance system which supports human operators, by automatically detecting loitering people. Usually, loitering human behavior often leads to abnormal situations, like suspected drug-dealing activity, bank robbery, and pickpocket, etc. Thus, the problem of loitering detection in image sequences involving situations with multiple objects is studied based two dimensional Markov random walks in which both motion and appearance features describing the movements of a varying number of objects as well as their entries and exits are used. To obtain efficient and compact representations we encode the spatiotemporal information of intra-inter trajectory contexts into the transition matrix of a Markov Random Walk, and then extract its stationary distribution and boundary crossing probabilities as final detection criteria. The model is also made less sensitive to uninteresting objects occluding the region of interest by integration out their effect on the observation probabilities. The resulting system is tested on the real life dataset scenarios giving 95% performance results. © 2010 IEEE.
-
A stochastic model for popularity measures in web dynamics Reviewed
H. Hama H., Pyke Tin, Thi Thi Zin, T. Toriu
Proceedings - 2010 6th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIHMSP 2010 676 - 679 2010.12
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Proceedings - 2010 6th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIHMSP 2010
In this paper, we propose a stochastic web dynamic model based on the concept of queuing theory to measure popularity of websites in the World Wide Web. We assume that the characteristics of a website such as novelty, popularity, reliability, and relevancy are governed by two major forces: internal or self-growth of each website and external functions acting on the website. In stochastic language, these two forces can be considered as random variables and able to investigate the important characteristics of websites. These characteristics include a probability distribution of the number of visitors to websites, the visiting times distribution, users' attention, the functional growth and decay of individual websites, the relationship between the dynamics of the web and its structure. We then define and investigate measuring process of popularity of websites using stochastic difference equations based on the structure of the web taken users' attention into account. For validation, we present some simulation results with the respect to parameter variations in the model. It shows that the proposed model can efficiently and adequately analyze the characteristics and behaviors of websites in web dynamical system. © 2010 IEEE.
-
HOG embedded Markov chain model for pedestrian detection
Zin T., Hama H., Tin P., Toriu T.
ICIC Express Letters 4 ( 6 B ) 2463 - 2468 2010.12
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters
This paper presents a new method for pedestrian detection by establishing Histograms of Oriented Gradients (HOG) embedded Markov chain model based on the cooccurrence of dominant orientations of gradients. In this model, HOG are used to obtain dominant orientations and their co-occurrences are computed by using various positional Structuring Elements (SEs). We then embed the HOG pair co-occurrences into a Markov chain. Having defined the embedded Markov chain, the corresponding joint probability density functions (pdf) are derived and used as Markov similarity features for pedestrian detection process. Due to the use of various positional SEs, the derived features can express complex shapes of objects with local and global distributions of gradient orientations. Experimental results on common datasets and comparison with some previous methods are given. The results show that the performance of our method is significantly well and it outperforms some existing methods such as conventional co-occurrence and HOGs. ICIC International © 2010 ISSN 1881-803X.
-
Multivariate web information reliability search engine
Hama H., Tin P., Zin T., Toriu T.
ICIC Express Letters 4 ( 6 B ) 2457 - 2462 2010.12
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters
This paper proposes a novel multivariate analysis approach to web information reliability search systems. In this analysis, we introduce a new representation of the web as a directed stochastic hyper-graph, instead of a simple graph, where links can connect not only pairs of web pages, but also pairs of disjoint sets of pages. The stochastic hyper-graph is constructed by regrouping the set of pages into non-overlapping page-sets (subsets) and using the links between pages of distinct page-sets to create weighted hyperarcs with the goal of providing more reliable information. We then embed the hyper-graph structure into a Layered Markov Model in which transitions among web sites, page-sets and web pages are distinguished to compute reliability of web information. In addition, personalized rankings which are keys to next generation search engine will be produced by adapting the computation at local and global layers. Finally, we present some illustrative simulation results showing that the ranking system generated by the proposed approach is qualitatively comparable to or even better and more reliable than the ranking produced by some famous search engines such as Google. ICIC International (5)2010 ISSN 1881-803X.
-
A novel way of computing multimedia information similarities
Zin T., Tin P., Toriu T., Hama H.
ICIC Express Letters, Part B: Applications 1 ( 1 ) 27 - 32 2010.9
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
This paper presents a new perspective on characterizing the similarity be- tween elements in multimedia information systems concerning with images, text docu- ments or, more generally, nodes of a weighted and undirected graph. It is based on a sequence of Embedded Markov-chains and its rich properties. More precisely, we com- pute quantities such as the steady state probability distributions and the average commute time, that provide similarities between any pair of nodes or information. This approach is not limited to only images but it is applicable to any type of data such as range sensor data, communication frequency data and so on. Besides, it is adaptable for detecting and retrieving information from the multimedia database by using our newly developed Markov-based similarity measures. Specifically, the proposed system contributes a new method of foreground/background modeling in visible or non-visible multi-sensors infor- mation processing, detection, tracking and analyzing human-machine related surveillance systems. Experimental results on real life data sets show that the Markov Chain-based similarities perform well in comparison with other methods. © 2010 ICIC International.
-
A stochastic model for web reliability ranking system Reviewed
Pyke Tin, Thi Thi Zin, T. Toriu, H. Hama
ICIC Express Letters 4 ( 3 ) 705 - 711 2010.6
Language：English Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters
The World Wide Web is used, by an increasing number of people as an ever expanding source of information on almost every topic imaginable. However, useful data, is often buried, in large quantities of low-reliability content. Estimation of Web information reliability is valuable for diverse applications, such as a, search result ranking and, a, direction of crawlers. In this paper, we propose a, novel stochastic model for web information ranking system, which enables us to search useful and, reliable knowledge information. Specifically, a multiple state reliability ranking model based on the theory of Markov chain is developed, by assuming that the likelihood, of a, statement on the Web can be trusted, using standards developed, by information scientists, and, the link structure of associated, web pages. We then cluster relevant and, reliable webpages based, on whether they can be trusted, or not. Finally, the proposed model is tested, on an academic search engine and, show how the reliability ranks can be used, for searching a, useful knowledge.
-
Human behaviors analysis at or near public transportation asset
Thi Thi Zin, K. Fujimura, S. Kamijo
17th ITS World Congress 2010
Authorship：Lead author Language：English Publishing type：Research paper (scientific journal) Publisher：17th ITS World Congress
Security of human lives and property has always been a major concern for civilization for several centuries. In modern civilization, the threats of theft, accidents, terrorists' attacks and riots are ever increasing in public access areas such as airports, train stations, shopping malls, banks and etc. Due to the high amount of useful information that can be extracted from a video sequence, video surveillance has come up as an effective tool to forestall these security problems. In surveillance systems, understanding human behaviors and activities arising out of the interactions of various objects, as well as their evolution over time is an important problem. The protection of critical transportation assets and infrastructure is an important topic in these days. In this paper, we develop a new rule based approach to smart video surveillance system for detecting situations where people may be in peril, as well as suspicious action or interactions at or near critical transportation assets. For organizational purposes, the surveillance operationally-relevant behaviors are divided into three general groups: (i) single person or no interaction, (ii) multiple person interactions, and (iii) person-facility/location interactions. The behavior analysis is accomplished through the development of geometric and motion visual features for each pedestrian. With this information, the system could alert authorities if pedestrians display suspicious behaviors. The performance evaluation of the proposed system is carried out by using the video sequences taken in the real life environments of rail stations. The experimental results show the high accuracy rates.
-
Optimal crawling strategies for multimedia search engines Reviewed
H. Hama, Thi Thi Zin, Pyke Tin
IIH-MSP 2009 - 2009 5th International Conference on Intelligent Information Hiding and Multimedia Signal Processing 182 - 185 2009.12
Language：English Publishing type：Research paper (international conference proceedings) Publisher：IIH-MSP 2009 - 2009 5th International Conference on Intelligent Information Hiding and Multimedia Signal Processing
In this paper we propose a novel optimal crawling strategy for next-generation multimedia search engines. We consider here a Web crawl as a two-dimensional (2D) random walker on a graph whose vertices are the Web pages and whose edges are the hyperlinks. The proposed crawler is a two-part scheme optimizing the crawling process in such a way that the average level of staleness over all pages is minimized and the quality of search engine from user's perspective is maximized. In doing so, we employ techniques from probability theory and the theory of functional equations which are highly computationally efficient-crucial for practicality because the size of the problem in the Web environment is immense. We show that a combination of breadth-depth crawling including the largest sites is a practical and optimal strategy. In particular, several probabilistic models for user browsing in infinite Web are proposed and studied to estimate how deep and breadth a crawler must go to download a significant portion of the Web site that is actually visited. Experimental and simulation results show that a crawler needs to download just a few levels in depth and breadth to reach the maximum number of pages that users actually visit. It also suggests that the largest sites should be included in the crawling process. © 2009 IEEE.
-
Dominant color embedded Markov chain model for object image retrieval
Zin T., Tin P., Toriu T., Hama H.
IIH-MSP 2009 - 2009 5th International Conference on Intelligent Information Hiding and Multimedia Signal Processing 186 - 189 2009.12
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：IIH-MSP 2009 - 2009 5th International Conference on Intelligent Information Hiding and Multimedia Signal Processing
This paper proposes a new and compact method for object image retrieval fusing Dominant Colors (DCs) and embedded Markov chain concepts. This proposed method uses combined color-texture features which are characterized in terms of their spatial interaction or interrelationship properties, modeled by means of a set of embedded Markov chains, each associated with a major spatial direction. Specifically, DCs are extracted from the object image, which are encountered pixel-wise along a given direction to form an embedded Markov chain. Normalizing the resultant Markov chains over all specified directions, the corresponding stationary distribution is derived and served as Markov Feature-Vector (MFV). We then employ the chi square distance between the feature vectors in comparing similarity of images. The MFV involves spatial structure information of both within and between dominant color regions. Moreover, it keeps simplicity, compactness, efficiency, and robustness. We conduct experiments using a comprehensive set of images including deformable shapes. Experimental results show that the proposed method can retrieve an important number of correct images with very high accuracy while the mismatch ratio remains constant. © 2009 IEEE.
-
Bundling multislit-HOG features of near infrared images for pedestrian detection
Zin T., Tin P., Hama H.
2009 4th International Conference on Innovative Computing, Information and Control, ICICIC 2009 302 - 305 2009.12
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：2009 4th International Conference on Innovative Computing, Information and Control, ICICIC 2009
In this paper we present a novel scheme where image features are bundled into local groups. Specifically, features of Near Infrared (NIR) images extracted by using Histogram of Oriented Gradients (HOG) descriptor and those by our multislit method are bundled into a single descriptor. The method involves first localizing the spatial layout of body parts (head, torso, and legs) in individual frames using multislit structures, and associating these through a series of extracting HOG features. A bundled feature vector describing various types of poses is then constructed and used for detecting the pedestrians. Experiments with a database of NIR images show that our scheme achieves a substantial improvement in average precision over the baseline conventional HOG approach. Detection and recognition performance is less computationally expensive than existing approaches. © 2009 IEEE.
-
An optimal choice of morphological operating center for object image retrieval
Hama H., Zin T., Tin P.
2009 4th International Conference on Innovative Computing, Information and Control, ICICIC 2009 298 - 301 2009.12
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：2009 4th International Conference on Innovative Computing, Information and Control, ICICIC 2009
In this paper we introduce a novel and simple schemes to develop an optimal choice of morphological Operating Center (OC) for object image retrieval. A variation of the standard morphological operators which require the choice of an OC is discussed. The proposed method is based on combinations of statistical and dynamic programming techniques in which recursive equations on the basis of dilation by using the principle of optimality and minimizing unnecessary background area of the objects are applied. We also present the application of mathematical morphology with Structuring Elements (SEs) which are elongated in the angular direction. The experimental results show that the optimal choice of OC provides satisfying retrieval results. © 2009 IEEE.
-
Reliability based web information ranking system
Tin P., Zin T., Toriu T., Hama H.
2009 4th International Conference on Innovative Computing, Information and Control, ICICIC 2009 294 - 297 2009.12
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：2009 4th International Conference on Innovative Computing, Information and Control, ICICIC 2009
In this paper, we propose a reliability based Web information ranking system which enables searching useful and reliable knowledge information. The proposed system will contain subsystems for reliability ranking, information clustering based on reliability. The reliability ranking system will estimate the likelihood that a statement on the Web can be trusted using standards developed by information scientists, and the link structure of associated Web pages. The clustering will cluster relevant and reliable information based on whether or not they can be trusted or not. We test these models on an academic search engine and show how the reliability information ranks can be used as a useful knowledge. © 2009 IEEE.
-
Spatial image retrieval based on dynamic thresholding
Zin T., Hama H., Tin P.
International Journal of Innovative Computing, Information and Control 5 ( 11 ) 4051 - 4059 2009.11
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：International Journal of Innovative Computing, Information and Control
In this paper, we present a spatial image retrieval method based on dynamic thresholding. The proposed method can retrieve spatial image patterns with high accuracy and speed from images with complicated backgrounds. For simplicity, we consider the query images as specified rectangular-shaped or circular-shaped framed images. First, by introducing a dynamic thresholding system, the images can be partitioned into Peak Color Regions (PCRs). Consequently, the proposed method requires low computational complexity giving optimal feasible results for detection and segmentation. Due to compact representation and low complexity of color features, direct histogram comparison is to be used for extraction of PCRs. Since the number of the PCRs is much smaller than that of the image pixels, the proposed method allows a low dimensional image processing. The effectiveness of the proposed method is confirmed through experiments with various images. © 2009 ISSN.
-
A hybrid ranking of link and popularity for novel search engine Reviewed
H. Hama, Thi Thi Zin, Pyke Tin
International Journal of Innovative Computing, Information and Control 5 ( 11 ) 4041 - 4049 2009.11
Language：English Publishing type：Research paper (scientific journal) Publisher：International Journal of Innovative Computing, Information and Control
In this paper, we explore a new paradigm to enable Web search at image level reflecting the most relevant results to the users. So we introduce a new concept of Popularity-based Rank (PR), a content-level ranking and searching model for image retrieval. Specifically, we establish a PR Operation which is combined with new link structure analysis. The strategy to decide searching order by taking similarity into consideration is also proposed and proved to be effective and efficient. Experimental results show that the combined analysis can achieve significantly better ranking results than naively applying page-level ranking on the image model which is usually used in the current search engines. © 2009 ISSN.
-
Zin T., Takahashi H., Hama H.
International Journal of Innovative Computing, Information and Control 5 ( 3 ) 751 - 761 2009.3
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：International Journal of Innovative Computing, Information and Control
Nowadays, person detection in far infrared (FIR) images toward realizing a night vision system becomes a hot topic. However, sufficient performance could not be achieved by conventional schemes. Since the properties of FIR images different from visible images, it is not known what kind of scheme is appropriate for person detection in FIR images. In this paper, we propose two novel methods for person detection using FIR images: (i) body parts detection method and (ii) Gravity Center (GC) movement pattern method. First, we introduce the multi-slit method along with vanishing line for extraction of head regions. After the head region is detected and segmented, the person body and legs regions are roughly estimated by size ratios. The histograms of Sobel edge of such estimated regions are used to confirm the segmented head. This method can be applicable to person detection at both near and far distances in indoor and outdoor scenes. Second, we propose a sequential decision method by investigating GC movement patterns. It is very simple and especially valid for images at near distances. Our experiments demonstrate the effectiveness of the proposed methods and the advantages in dealing with person detection. Finally, comparative study and further extendable potential applications of the proposed methods are pointed out to be focused in our future research. © 2009 ISSN.
-
Ranking system for image database using special type of markov chain
Hama H.
Proceedings - Digital Image Computing: Techniques and Applications, DICTA 2008 556 - 563 2008.12
Language：English Publishing type：Research paper (scientific journal) Publisher：Proceedings - Digital Image Computing: Techniques and Applications, DICTA 2008
In this paper, we explore and examine a new novel approach to image ranking systems based on some special types of Markov chain along with new concepts of popularity and relevancy measures for image database. To be specific, this approach introduces a family of special Markov chain models in which serial correlations are explicitly involved so that we can use them as correlations among the images. By using these models, we develop a ranking function for the image database. On the other hand, 'popularity' and 'relevancy' concepts are introduced and used for developing an alternative ranking function for the database. In the process of developing ranking functions we use a method of queue-based stochastic difference equations. We then blend two ranking functions, to propose a new ranking system for searching order of image database. Since the proposed ranking system considers concepts of correlations, popularity and relevancy altogether, it is beneficial to a modern search engine for investigating behavior and effects of those parameters on the search results. Some illustrative examples and simulation results are presented with reference to a real world application domain. © 2008 IEEE.
-
Zin T., Hama H., Tin P.
Proceedings - Digital Image Computing: Techniques and Applications, DICTA 2008 548 - 555 2008.12
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Proceedings - Digital Image Computing: Techniques and Applications, DICTA 2008
This paper proposes a new method for object retrieval in image and video databases. The proposed system uses the histogram based approach along with Angular Radial Representation (ARR). In addition, concepts of Dominant Colors (DCs) and morphological dilation using ring-shaped and fanshaped Structuring Elements (SEs) are also applied. It is found that the approach is not only invariant to rotation, translation and scaling but also valid for low resolution images and partial occlusion. The retrieval effectiveness of the proposed system is shown through experiments using a comprehensive set of images including deformable shapes. © 2008 IEEE.
-
Novel search engine: Combination of link and popularity rank for multimedia retrieval
Hama H., Zin T., Tin P.
3rd International Conference on Innovative Computing Information and Control, ICICIC'08 2008.9
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：3rd International Conference on Innovative Computing Information and Control, ICICIC'08
In contrast with the current search engines that essentially do page-level ranking and searching, we are exploring a new paradigm to enable Web search at image level by introducing a new concept of Popularity-based Rank (PR). This paper introduces PR, a content-level ranking and searching model for multimedia retrieval. Specifically, we establish a PR Operation which is combined with new link structure analysis. The strategy to decide searching order by taking similarity into consideration is also proposed and proved to be effective and efficient. Experimental results show that the combined analysis can achieve significantly better ranking results than naively applying PageRank on the image model. © 2008 IEEE.
-
Robust detection and segmentation of images with tolerance
Zin T., Hama H., Tin P.
3rd International Conference on Innovative Computing Information and Control, ICICIC'08 2008.9
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：3rd International Conference on Innovative Computing Information and Control, ICICIC'08
In this paper, we introduce a novel approach to retrieve significant spatial images from a collection of images. The query images of interest in this paper are specified rectangular-shaped and circular-shaped as framed photographs. By incorporating the advantages of adaptive thresholding and tolerant partitioning into peak color regions, the proposed method requires low computational complexity and it is very feasible for image detection and segmentation. Peak image regions are then extracted by using color histograms. Since the number of the peak regions is much smaller than that of the image pixels, this proposed method allows a low dimensional image processing. The effectiveness of the proposed method is confirmed through experiments with various images. © 2008 IEEE.
-
Robust person detection using far infrared camera for image fusion Reviewed
Thi Thi Zin, H. Takahashi, H. Hama
Second International Conference on Innovative Computing, Information and Control, ICICIC 2007 2008.2
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Second International Conference on Innovative Computing, Information and Control, ICICIC 2007
In this paper we present a robust method for person detection using far infrared images. To extract initial nominated head regions, thresholding and morphological operations are applied using intensity information. Among these regions, some of wrongly extracted regions are removed using the pattern of person head based on the local maximums of Sobel edge image. After the head regions are segmented, the person body and legs region are roughly estimated by the ratios. The histograms of Sobel edge of such estimated regions are used to confirm the segmented head. This method can be applicable to person detection at both near and far distances in indoor and outdoor scenes. Moreover, we propose another novel algorithm using the movement pattern of gravity centers. It is a very simple way, especially valid for images at near distances. Our experiments demonstrate the effectiveness of the proposed method and the advantages in dealing with person detection for night vision applications. Finally, image fusion of visible and far infrared cameras is discussed.
-
Robust signboard recogniton in the presence of occlusion and reflection Reviewed
Thi Thi Zin, H. Hama, S-S. Koh
International Journal of Innovative Computing, Information and Control 3 ( 6 A ) 1321 - 1334 2007.12
Language：English Publishing type：Research paper (scientific journal) Publisher：International Journal of Innovative Computing, Information and Control
Recognizing objects for visual information in outdoor scenes is very useful but challenging. This paper presents a new framework to recognize signboards with uniform color regions in the presence of occlusion and reflection. The framework is composed of three stages: (i) extraction of uniform color regions by adaptive rank filter and piecewise linear approximation, (ii) recognition by template matching, and (iii) verification by relative color polygons. In the experiments, we used 300 images taken under a great variety of adverse conditions including occlusion, reflection, specular highlights, and so on. The proposed system achieved 98% recognition rate for images taken under such bad conditions. Moreover, it can be extended for vision-based car and pedestrian navigation systems to provide up-to-date information to the users and potentially be embedded in a driver assistance system.
-
Accurate reconstruction of non-rigid 3D shapes Reviewed
S-S. Koh, Thi Thi Zin, H. Hama
Digest of Technical Papers - IEEE International Conference on Consumer Electronics 2007.8
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Digest of Technical Papers - IEEE International Conference on Consumer Electronics
The aim of this paper is to create an accurate 3D-TV contents from 2D-video which can be used in a commercial 3D-TV broadcast environment. In this paper we propose an accurate recovering of non-rigid 3D shapes using a chained partial rigid factorization method. We have extended the existing rigid factorization algorithm to the stereo camera case. Our focus in this paper is on estimating deformable shapes from stereo images. Our experiments show the improved results for reconstructions. ©2007 IEEE.
-
Robust signboard recognition for vision-based navigation Reviewed
Thi Thi Zin, S-S. Koh, H. Hama
Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers 61 ( 8 ) 1192 - 1200 2007.8
Language：English Publishing type：Research paper (scientific journal) Publisher：Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers
We propose a framework that enables vision-based navigation systems to more robustly recognize signboards. It is designed mainly for signboards with uniform color regions, which can be strongly affected by occlusion and reflection. The framework comprises two stages: preprocessing for extracting uniform color regions and recognizing them. In the first stage, the cumulative frequency, morphology gradient, and Mahalanobis distance are used for the extraction. In the second stage, a new matching method for identifying templates and a new relative method for verifying colors are used. Testing of a system using this framework and 300 images taken under various conditions showed a recognition rate of 98%, demonstrating the framework's powerful ability to recognize signboards.
DOI： 10.3169/itej.61.1192
-
Optimal color space for relative color polygons
Zin T., Koh S., Hama H.
IEICE Electronics Express 4 ( 3 ) 106 - 113 2007.3
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：IEICE Electronics Express
We aim for detection and recognition of planar objects in natural outdoor scenes under varying illumination conditions. To achieve this, we propose Relative Color Polygons (RCPs) using component colors of objects for color matching. They can be defined on many color spaces, and it is found that a 2D color space (XY space) is the optimal color space for our relative color method compared with other color spaces. To evaluate the invariance to illumination changes for object recognition, experiments have been carried out using 500 outdoor scene images. By using the proposed model, the color matching rate of the input images with the standard one was 95%. This framework is potentially applicable to image retrieval, image segmentation, image recognition, and so on. © IEICE 2007.
DOI： 10.1587/elex.4.106
-
Geometrical properties between 2-D image plane and 3-D error space
Koh S., Zin T., Hama H.
IEICE Electronics Express 3 ( 14 ) 333 - 339 2006.7
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：IEICE Electronics Express
We study geometrical properties between 2-D image plane and 3-D error space under affine reconstruction. The purpose of our system is to contribute to more accurate 3-D reconstruction by analyzing geometrically 2-D to 3-D relationship. In situation for no missing feature points and no noise in the 2-D observation matrix, the accurate solution is known to be provided by Singular Value Decomposition. However, several feature points of the matrix have not been observed because of occlusions and low image resolution, and so on. In this case, there is no simple solution. To obtain accurate 3-D reconstruction by recovering missing feature points, we propose the analytic approach which can handle the error orientation and distance of missing feature points by the geometrical properties. © IEICE 2006.
DOI： 10.1587/elex.3.333
-
Accurate estimation of missing data under noise distribution Reviewed
S-S. Koh, Thi Thi Zin, H. Hama
IEEE Transactions on Consumer Electronics 52 ( 2 ) 528 - 535 2006.5
Language：English Publishing type：Research paper (scientific journal) Publisher：IEEE Transactions on Consumer Electronics
3D contents have been becoming one of attractive multimedia and reality. In computer vision, 2D-to-3D conversion techniques require estimating missing data from noisy observations. When there is no missing data in the observation matrix, the accurate solution of such problem is known to be given by Singular Value Decomposition (SVD). In the case of converting already recorded monoscopic video contents to 3D, several entries of the matrix have not been observed. Therefore, the problem has no simple solution, so it is necessary to estimate missing data. In this paper, we propose an estimation algorithm of missing data with minimizing the influence of noise embedded when tracking feature points from partial observations. The proposed method is an iterative affine SVD factorization method which can estimate the model parameters, given an incomplete set of the observation matrix. The main idea of our algorithm is to estimate missing data accurately even under noise distribution by using geometrical correlations between 2D and 3D error space. This paper consists of three main phases: geometrical correlations for estimating missing data, estimation algorithm, and analyzing the results for video sequences. The accurate results in practical situations as demonstrated here with synthetic and real video sequences show the efficiency and flexibility of the proposed method.
-
Relative color polygons for object detection and recognition
Zin T., Koh S., Hama H.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 4259 LNAI 834 - 843 2006
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
This paper proposes a new framework of the color model for outdoor scene image detection and recognition. This model enables us to manipulate easily the color of an image. Here, the concept of 'relative color polygon' for an object composed of uniform color regions is introduced on a 2D color space (XY space). Then the color similarity is defined using three kinds of parameters of the polygon: length and slope of every side and angle of adjacent sides. This paper addresses how to decide the color similarity by using the facts about color shifting on the XY space. The feasibility of the proposed framework has been confirmed through the experimental results using outdoor scene images taken under a great variety of various illumination conditions. © Springer-Verlag Berlin Heidelberg 2006.
-
Robust segmentation of road traffic signs using adaptive thresholds
Liao C.
IEICE Electronics Express 2 ( 14 ) 423 - 428 2005.11
Language：English Publishing type：Research paper (scientific journal) Publisher：IEICE Electronics Express
We propose a robust segmentation system for road traffic signs using adaptive thresholds. The main objective is to develop a robust segmentation system under various illumination conditions. The main problem is how to obtain the best suited threshold for relative similarity. It can be got automatically by K-nearest connectivity when K = 4 and the two-way connectivity method. Our system utilizes relative similarity of neighborhood pixels and shape information of traffic signs together. The effectiveness of our proposed system is confirmed through experiments under various illumination conditions. In the ex-periments, 90 images taken under daytime, evening and night-time are used. The system can give 98% successful segmentation rate. © 2005, The Institute of Electronics, Information and Communication Engineers. All rights reserved.
DOI： 10.1587/elex.2.423
-
A robust road sign recognition using segmentation with morphology and relative color Reviewed
Thi Thi Zin, H. Hama
映像情報メディア学会 59 ( 9 ) 1333 - 1342 2005.9
Language：English Publishing type：Research paper (scientific journal) Publisher：Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers
We propose a robust road sign recognition system under various illumination conditions. The proposed approach has two steps: segmentation and recognition. The segmentation, which is the focus of this paper, is performed using morphological operations and relative color. The segmented regions are recognized by a template matching method using modified standard deviation. The algorithm works for various types of circular and pentagonal road signs. In experiments under various illumination conditions, the segmentation rate was 100% in the daytime and evening and 80% even in the night-time, and the recognition rate was 100% for all of the segmented regions under all illumination conditions. The effectiveness of the proposed system was confirmed through experiments using 200 images of road signs taken under a great variety of illumination conditions including fog and light rain.
-
Robust road sign recognition using standard deviation
Zin T., Hama H.
IEEE Conference on Intelligent Transportation Systems, Proceedings, ITSC 429 - 434 2004.12
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：IEEE Conference on Intelligent Transportation Systems, Proceedings, ITSC
This paper proposes the recognition of road signs under various illumination conditions using standard deviation. The proposed road sign recognition system uses both of shape and color information. The former one is used in template matching with standard deviation and the latter one in modified Mahalanobis distance. This paper focuses on circular road signs and achieves almost perfect performances under various conditions such as daytime, evening-time and nighttime. Moreover, the proposed method may be applicable to recognize of any shape of road signs. The effectiveness of the proposed method has been demonstrated through experiments. According to our experimental results, the recognition rate is 100% for 200 images taken under various illumination conditions.