Presentations -
-
Application of Methods in Sequential Analysis to Dairy Cow Calving Events International conference
Thi Thi Zin, K. Sumi, Pann Thinzar Seint, Pyke Tin, I. Kobayashi, Y. Horii
15th International Conference on Innovative Computing, Information and Control (ICICIC2021) (Online) 2021.9.15 ICIC International
Event date： 2021.9.15 - 2021.9.16
Language：English Presentation type：Oral presentation (general)
Venue：Online
Apart from sequential sampling, methods in the sequential analysis have been widely and successfully used for various applications such as insurance problems, theory of storage, queuing theory, and many other stochastic models. Moreover, it is well recognized that Wald’s Fundamental Identity in sequential analysis can be used to derive approximate and some exact results in most situations wherein we have essentially a random sequence phenomenon. In this aspect, the fluctuations in motion of a pregnant cow around the calving event fall into a random sequence category. Therefore, in this paper, we explore and examine an application of Wald’s Fundamental Identity in sequential analysis to dairy cow calving time prediction models. Specifically, we will show this Fundamental Identity which can be used to derive results for predicted calving times at which an individual cow calving event occurs in a video-monitored maternity barn. The paper is pure of an expository nature and considers only simple illustrations and some real-life video data are used to confirm the proposed method.
-
Cattle Region Extraction Using Color Space Conversion International conference
Y. Hashimoto, H. Hama, Thi Thi Zin
15th International Conference on Innovative Computing, Information and Control (ICICIC2021) (Online) 2021.9.15 ICIC International
Event date： 2021.9.15 - 2021.9.16
Language：English Presentation type：Oral presentation (general)
Venue：Online
The Japanese livestock industry has problems such as difficulty of finding successors of farms, and the aging farmers. Therefore, development of a cattle monitoring system through non-contact and non-invasive methods to improve productivity and reduce labor burdens is a strong desire from farmers. As one of elemental technologies to realize it, we have focused on tracking of cattle for detecting characteristic behaviors using video camera. In this paper, we present an effective extracting method of the cattle region from video images of pasture. Inter-frame difference and background subtraction are widely used for detecting moving objects in video images; however, in our case the system is supposed to be used in dusty pasture, and they do not work well. Because Japanese black cattle move slowly in general, and the skin color is similar to soil, region extraction of them is very difficult, even if these conventional methods are used. Here, region extraction using Color Space Conversion is adopted. This method enables us to manipulate easily the image’s color and automatically extract cattle region; additionally, the movement of cattle is estimated from the change of the gravity center of the extracted region. To verify the effectiveness, we carried out an experiment using the videos taken at Sumiyoshi Livestock Science Station at University of Miyazaki. The experimental results show that this method is well suited for extracting cattle region. Further verification should be conducted to enhance robustness.
-
Cow Region Segmentation in Cattle Farm by Using Semantic Segmentation Networks International conference
Swe Zar Maw, Thi Thi Zin, Pyke Tin
15th International Conference on Innovative Computing, Information and Control (ICICIC2021) (Online) 2021.9.15 ICIC International
Event date： 2021.9.15 - 2021.9.16
Language：English Presentation type：Oral presentation (general)
Venue：Online
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 cows. 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 module 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.
-
A Study on Working Group Detection after Helmet Extraction International conference
S. Inoue, I. Hidaka, Thi Thi Zin
15th International Conference on Innovative Computing, Information and Control (ICICIC2021) (Online) 2021.9.15 ICIC International
Event date： 2021.9.15 - 2021.9.16
Language：English Presentation type：Oral presentation (general)
Venue：Online
Due to the declining birthrate and increasing in aging population, the industries are facing seriously with manpower shortage problems. As a consequence, many small and medium-sized factories are introducing AI and IOT to automate their operations so that they can cope with fewer employees. However, some factories are not automating because it is more economically effective to have employees perform the work directly than to automate it. Such companies can create more economic benefits with less labor by eliminating waste from their current work. In this research, we focus on the movement of people during work by using information obtained from a 4K camera installed overhead that can detect work groups and acquire data for use in creating indicators for efficiency improvement. For this purpose, we attach a marker on the top of the worker’s helmet to detect the helmet and identify the work group. The effectiveness of the proposed method has been confirmed through simulation experiments.
-
Holistically-Nested Deep Learning Model for Cow Region Detection and Motion Classification International conference
Thi Thi Zin, Saw Zay Maung Maung and Pyke Tin
15th International Conference on Innovative Computing, Information and Control (ICICIC2021) (Online) 2021.9.15 ICIC International
Event date： 2021.9.15 - 2021.9.16
Language：English Presentation type：Oral presentation (general)
Venue：Online
In dairy farming, monitoring an individual cow is a critical component for each and every aspect of precision dairy farm management system since it can make farmers and managers learn cow’s health conditions, body conditions even the occurrences of calving difficulties in times. Such monitoring is often performed visually because animal appearance and behavior are key indicators of analyzing animal conditions. According to the latest computer vision and image processing algorithms, it is now possible to implement a monitoring system that can detect the status of cows at a low cost. In this aspect, one of the important steps is to detect and segment the cows within the video sequences. Moreover, how an individual cow behaves or makes movements such as lying, standing, and changes from one posture to another is also equally important for further analysis. Therefore, in this paper, we propose a deep learning method of edge-based cow region detection and multiple linear models for the classification of cow movements. Specifically, we establish a deep learning model of holistically-nested edge detection (HED) that performs image-to-image prediction by using fully convolutional neural networks and deeply supervised nets. In the cow motion classification process, we propose multiple linear models in which the coefficients of independent variables are utilized as features for classification. According to our experimental results, the proposed detection system is promising and provides robust performance. Similar experiments are also performed to validate the proposed multiple linear models for the classification of cow motions with high accuracy.
-
Automatic Detection of Mounting Behavior in Cattle using Semantic Segmentation and Classification International conference
Su Myat Noe, Thi Thi Zin, Ikuo Kobayashi, Pyke Tin
The 2021 IEEE 3rd Global Conference on Life Sciences and Technologies (LifeTech 2021) (Nara Royal Hotel (Nara, Japan)) IEEE Life Sciences Technical Community
Event date： 2021.3.9 - 2021.3.11
Language：English Presentation type：Oral presentation (general)
Venue：Nara Royal Hotel (Nara, Japan)
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 International conference
Htin Lynn Aung, Thaire Phyu Win, Thi Thi Zin
The 2021 IEEE 3rd Global Conference on Life Sciences and Technologies (LifeTech 2021) (Nara Royal Hotel (Nara, Japan)) IEEE Life Sciences Technical Community
Event date： 2021.3.9 - 2021.3.11
Language：English Presentation type：Oral presentation (general)
Venue：Nara Royal Hotel (Nara, Japan)
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 International conference
Thi Thi Zin, Pyke Tin, Pann Thinzar Seint, Yoichiro Horii
The 2021 IEEE 3rd Global Conference on Life Sciences and Technologies (LifeTech 2021) (Nara Royal Hotel (Nara, Japan)) IEEE Life Sciences Technical Community
Event date： 2021.3.9 - 2021.3.11
Language：English Presentation type：Oral presentation (general)
Venue：Nara Royal Hotel (Nara, Japan)
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.
-
Image Technology Based Detection of Infected Shrimp in Adverse Environments International conference
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 (Prince of Songkla University) www.psu.ac.th, www.kku.ac.th, www.ku.ac.th, www.cmu.ac.th,
Event date： 2021.1.11 - 2021.1.12
Language：English Presentation type：Oral presentation (general)
Venue：Prince of Songkla University
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.
-
Consumer Behavior Analyzer in Internet of Things (IoT) Environments International conference
Swe Nwe Nwe Htun, Thi Thi Zin and Pyke Tin
4th International Symposium on Information and Knowledge Management (ISIKM2020) (Online Conference) ICIC International
Event date： 2020.12.12 - 2020.12.13
Language：English Presentation type：Oral presentation (general)
Venue：Online Conference
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.
-
Systematic Inclusion Study on Some Rare Gemstones of the Mogok Area, Mandalay Region, Myanmar International conference
Htin Lynn Aung, Thaire Phyu Win and Thi Thi Zin
4th International Symposium on Information and Knowledge Management (ISIKM2020) (Online Conference) ICIC International
Event date： 2020.12.12 - 2020.12.13
Language：English Presentation type：Oral presentation (general)
Venue：Online Conference
In this paper we shall explore and examine an inclusion aspect of some rare gemstones of Mogok area which is known as the Ruby Land of Myanmar where 90% of world rubies come from. The materials presented in this paper are the products of our research team investigated some rock sequences in the Mogok area situated about 280 miles north of the capital Naypyidaw, has unearthed some of the rarest and most luxurious rubies including the legendary 82-carat Nga Mauk ruby discovered centuries ago.. The rock sequence of the study area consists of medium to high grade metamorphic rocks, marble, gneiss and intrusive igneous rocks, mainly Kabaing granite, leucogranite and syenite. It is famous for presence of ruby and sapphire. Exceptionally some rare gemstones are also discovered. The present work is mainly intended to describe 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.
-
Smart Irrigation: An Intelligent System for Growing Strawberry Plants in Different Seasons of the Year International conference
Ye Htet, Htin Kyaw Oo and Thi Thi Zin
4th International Symposium on Information and Knowledge Management (ISIKM2020) (Online Conference) ICIC International
Event date： 2020.12.12 - 2020.12.13
Language：English Presentation type：Oral presentation (general)
Venue：Online Conference
Agriculture is an important source of livelihood and varying the way of cultivating plants could provide more productivity and sustainability of foods than before and thus smart irrigation would be one of the best solutions. Therefore, the proposed system mainly focused on the strawberry plants to grow within small-scale farm using intelligent systems to bear and produce fruits in all seasons of the country. A challenging problem which arises for this objective is the precise temperature, water and fertilizer management for plants. So, this system emphasized on automatic environmental adjustment system integrated with sensors to control temperature and also drip irrigation system for efficient water and fertilizers usage. Moreover, leaf analysis using computer vision which is controlled by Raspberry Pi is implemented for detection of the nutrient deficiency symptoms of plants. As for the communication unit to inform the users via sensors and image processing, Internet of Things is adopted.
-
Dam Water Overflow Estimation using Time Series International conference
Mie Mie Khin, Mie Mie Tin, Thi Thi Zin, Pyke Tin
2020 IEEE 9th Global Conference on Consumer Electronics (GCCE) (Kobe, Japan) IEEE
Event date： 2020.10.13 - 2020.10.16
Language：English Presentation type：Oral presentation (general)
Venue：Kobe, Japan
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.
-
Predicting Calving Time of Dairy Cows by Exponential Smoothing Models International conference
Tunn Cho Lwin, Thi Thi Zin, Pyke Tin
2020 IEEE 9th Global Conference on Consumer Electronics (GCCE) (Kobe, Japan) IEEE
Event date： 2020.10.13 - 2020.10.16
Language：English Presentation type：Oral presentation (general)
Venue：Kobe, Japan
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.
-
Detection of Estrus in Cattle by using Image Technology and Machine Learning Methods International conference
Su Myat Noe, Thi Thi Zin, Pyke Tin, Hiromitsu Hama
2020 IEEE 9th Global Conference on Consumer Electronics (GCCE) (Kobe, Japan) IEEE
Event date： 2020.10.13 - 2020.10.16
Language：English Presentation type：Oral presentation (general)
Venue：Kobe, Japan
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.
-
Image Processing and Statistical Analysis Approach to Predict Calving Time in Dairy Cows International conference
Swe Zar Maw, Thi Thi Zin, Pyke Tin
2020 IEEE 9th Global Conference on Consumer Electronics (GCCE) (Kobe, Japan) IEEE
Event date： 2020.10.13 - 2020.10.16
Language：English Presentation type：Oral presentation (general)
Venue：Kobe, Japan
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 International conference
Thi Thi Zin, Saw Zay Maung Maung, Pyke Tin, Yoichiro HORII
2020 IEEE 9th Global Conference on Consumer Electronics (GCCE) (Kobe, Japan) IEEE
Event date： 2020.10.13 - 2020.10.16
Language：English Presentation type：Oral presentation (general)
Venue：Kobe, Japan
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.
-
A Mobile Application for Offline Handwritten Character Recognition International conference
Thi Thi Zin, Moe Zet Pwint, Shin Thant
2020 IEEE 9th Global Conference on Consumer Electronics (GCCE) (Kobe, Japan) IEEE
Event date： 2020.10.13 - 2020.10.16
Language：English Presentation type：Oral presentation (general)
Venue：Kobe, Japan
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.
-
Elderly Monitoring and Action Recognition System Using Stereo Depth Camera International conference
Thi Thi Zin, Ye Htet, Yuya Akagi, Hiroki Tamura, Kazuhiro Kondo, Sanae Araki
2020 IEEE 9th Global Conference on Consumer Electronics (GCCE) (Kobe, Japan) IEEE
Event date： 2020.10.13 - 2020.10.16
Language：English Presentation type：Oral presentation (general)
Venue：Kobe, Japan
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.
-
画像処理技術を用いた牛のモニタリングシステム Invited
Thi Thi Zin, 小林 郁雄, 椎屋 和久, PYKE TIN, 堀井 洋一郎, 濱 裕光
電気・情報関係学会九州支部連合大会 （第７３回連合大会） (オンライン開催 （大会本部：九州産業大学）) 電気・情報関係学会九州支部連合大会委員会
Event date： 2020.9.26 - 2020.9.27
Language：Japanese Presentation type：Oral presentation (general)
Venue：オンライン開催 （大会本部：九州産業大学）
高齢化、大規模化する現代の畜産で、24時間 365日にわたり家畜の健康管理を適切に行い、異常や変化に注意し続けながら経営を継続することは容易ではない。本研究開発では、家畜生産性の向上と地域活性化の実現を目的とする牛のモニタリングシステム構築に必要な要素技術の開発を行う。ここでは、牛の体脂肪率の指標となる BCS（Body Condition Score)を、画像解析技術を用いて自動的に評価し、その経時変化から健康状態をモニタリングできるシステムに関する開発について発表する。