|
所属 |
工学教育研究部 工学科機械知能プログラム担当 |
|
職名 |
助教 |
|
連絡先 |
|
|
関連SDGs |
|
学位 【 表示 ／ 非表示 】
-
博士 （ 2021年9月 宮崎大学 ）
-
修士 （ 2017年9月 宮崎大学 ）
-
修士 （ 2018年8月 University Of Computer Studies, Mandalay ）
-
学士 （ 2014年2月 University Of Computer Studies, Mandalay ）
論文 【 表示 ／ 非表示 】
-
Kawasue K., Win K.D., Tokunaga T.
Animals 16 ( 3 ) 2026年2月
担当区分：責任著者 記述言語：英語 掲載種別：研究論文（学術雑誌） 出版者・発行元：Animals
Live weight is widely used as a reference indicator for growth performance and for evaluating the accuracy of weight measurement technologies in pig production. However, live weight is not a fixed physiological quantity, and finishing pigs naturally experience substantial short-term mass fluctuations due to normal behaviors such as drinking, feeding, urination, and defecation. In this study, we integrated published physiological and behavioral parameters into a stochastic simulation model to quantify within-day live-weight dynamics in finishing pigs weighing approximately 100 kg. The simulation was conducted with 1-min temporal resolution over a 24-h period. The model demonstrated that short-term weight fluctuations of approximately ±3–5 kg can occur within a single day, even when measurement error is minimal. Across 1000 simulated pigs, the mean daily fluctuation range was 4.2 kg, confirming that kilogram-scale variation is physiologically expected under normal conditions. These results provide a plausible physiological basis for understanding the frequently reported discrepancies between camera-based weight estimates and instantaneous floor-scale measurements. Camera systems primarily reflect body mass derived from external morphology, whereas floor scales measure instantaneous total mass that includes transient contributions from gastrointestinal contents, ingested water, and retained waste. Consequently, direct comparisons based on instantaneous scale readings can be misleading when used as ground truth. Our findings indicate that commonly cited accuracy claims of ±2–3 kg for camera weighing systems should be interpreted with caution, as normal physiological weight variation often exceeds this range. Recognizing live weight as a dynamic physiological variable is essential for developing biologically meaningful evaluation frameworks and for the appropriate interpretation and comparison of weight measurement technologies in precision livestock farming.
DOI： 10.3390/ani16030498
-
Image-based quantification of chicken carcass size using deep learning and image processing to support screening in poultry processing. 査読あり
Kodama T, Win KD, Kawasue K, Narita T
Ｔｈｅ Ｊｏｕｒｎａｌ ｏｆ Ｖｅｔｅｒｉｎａｒｙ Ｍｅｄｉｃａｌ Ｓｃｉｅｎｃｅ advpub ( 0 ) 2026年
記述言語：英語 掲載種別：研究論文（学術雑誌） 出版者・発行元：公益社団法人 日本獣医学会
Miyazaki Prefecture is one of Japan’s leading regions in chicken production, where strict poultry inspection is conducted at processing plants. Licensed veterinarians serving as certified poultry inspectors visually examine approximately 50,000 chicken carcasses per day, raising concerns about inspector burden and inspection reliability. This study developed an image-based method to support screening of nonconforming carcasses, focusing on emaciation and wooden breast (WB). Video footage of the processing line was geometrically corrected and both carcasses and their supporting shackles were detected using You Only Look Once version 8 (YOLOv8). The trained model achieved a validation mean average precision at 50% intersection over union (mAP50) of 0.995. The line speed was estimated to track carcasses across frames, and known shackle dimensions were used as a reference scale to calculate carcass area and an aspect ratio defined as a thickness index. The thickness index reflects torso fullness and complements carcass area as an independent feature. Visual inspection by a former poultry inspector indicated consistency between the calculated indices and visual impressions such as “emaciated” or “WB”. Scatterplots of area versus thickness index showed only a weak positive correlation, suggesting that these indices capture distinct aspects of carcass conformation. Applying threshold lines on the plots allowed classification of some nonconforming carcasses, demonstrating the potential utility of these parameters as supplementary inspection indicators. This approach may reduce inspector burden, standardize inspection quality, and contribute to the development of future automated poultry inspection systems.
DOI： 10.1292/jvms.25-0559
-
Win K.D., Kawasue K., Kaneko M.
Sensors 25 ( 6 ) 2025年3月
担当区分：筆頭著者 掲載種別：研究論文（学術雑誌） 出版者・発行元：Sensors
Highlights: What are the main findings? A non-contact method for measuring neonatal head circumference using a 3D imaging sensor was developed. What is the implication of the main finding? A top-view 3D imaging system can be used to estimate the head circumferences of neonates without direct contact, ensuring safety for their delicate skin. Because some portions of the head may be missing in the top-view images, these missing areas are estimated using mat information. The extracted features and head surface values contribute to the head circumference measurement. This approach enhances the feasibility of non-contact head circumference measurement, which could improve neonatal monitoring in clinical settings. In Japan, birth rates are declining, but there are a rising number of underweight newborns who require specialized care in neonatal intensive care units (NICUs). Head circumference is an important indicator of brain development for low-birth-weight infants. However, measuring head circumference requires extreme care because low-birth-weight infants have fragile skin. Therefore, a non-contact measurement system using a 3D imaging sensor was developed. Using this system, three-dimensional data for a newborn’s head can be obtained from outside the incubator. Briefly, the images are taken from above the incubator, so there is an area behind the head that cannot be captured by the camera, but the head circumference estimation takes into account the fact that the head is in contact with the mat. The proposed method allows head circumference estimation without touching the newborn. This approach minimizes stress for both the neonate and the nurse and improves efficiency and safety in the NICU.
DOI： 10.3390/s25061869
-
Robust pig extraction using ground base depth images for automatic weight estimation 査読あり
Win K.D., Kawasue K., Tokunaga T.
Artificial Life and Robotics 30 ( 1 ) 42 - 50 2025年2月
担当区分：筆頭著者 掲載種別：研究論文（学術雑誌） 出版者・発行元：Artificial Life and Robotics
Dark colored pigs (Berkshire, Duroc, etc.) are widely recognized nationwide in Japan for their exceptional taste, with the southern Kyushu region being a renowned production area for these esteemed breeds. However, estimating the weight of these pigs using a camera presents a unique challenge. The key process in a camera-based weight estimation system is the precise extraction of the target pig from the background. Typically, cameras capture images from above, as the top-view images provide the most specific growth indicators. However, the image from above contains a ground image. Since Berkshire and Duroc pigs are black and red, respectively, they blend into the ground, making it difficult to accurately segment the pigs in the images. Thus, it is crucial to perfectly distinguish between the ground and the pigs. Therefore, a new extraction method is proposed to distinguish between the ground and pigs by converting depth data based on the pig's position. To enhance the efficiency of pig farming and alleviate the burden on workers, our goal is to develop a system that automatically measures the weight of Berkshire pigs for shipment without background interference. In this study, we installed the system at a Berkshire pig farm and demonstrated the effectiveness of this innovative extraction method for camera-based weight estimation.
MISC 【 表示 ／ 非表示 】
-
バークシャー種肥育豚における3Dカメラを用いた体重推定 査読あり
泊鞠沙, 仲間ひなた, 浦川啓介, 川末紀功仁, WIN Khin Dagon, 徳永忠昭
日本暖地畜産学会報 67 ( 2 ) 2024年
掲載種別：速報，短報，研究ノート等（学術雑誌）
-
機械学習を用いた豚の体重推定 査読あり
金澤波音, 川末紀功仁, KHIN DAGON Win, 壱岐侑祐
計測自動制御学会システムインテグレーション部門講演会(CD-ROM) 22nd 2021年
掲載種別：速報，短報，研究ノート等（学術雑誌）
-
カメラを用いたオートソーティング装置による肥育豚省力出荷体系 査読あり
岐本博紀, 岩切正芳, 壱岐侑祐, 三角久志, 加地雅也, 島村勝則, 松窪敬介, 吉田久美子, 冷水忠, 日高良太, 合志文利, 川島知之, 川末紀功仁, WIN Khin Dagon
九州沖縄農業試験研究の成果情報(Web) 2020 2020年
掲載種別：速報，短報，研究ノート等（学術雑誌）
講演・口頭発表等 【 表示 ／ 非表示 】
-
Feasibility Study on Automated Screening of Substandard Broiler Chickens Using YOLOv8
Takafumi KODAMA, Tsubasa NARITA, Khin Dagon Winb and Kikuhito KAWASUE
International Joint Conference on Advanced Mechatronics 2025 2026年11月27日
開催年月日： 2026年11月27日
記述言語：英語 会議種別：口頭発表（一般）
-
RGB-D カメラと熱画像のセンサフュージョンによる 三次元温度測定手法の検討
福西 賢大郎、川末 紀功仁、加藤 一圭、Kin Dagon Win
日本機械学会 九州支部 第79期総会・講演会 2025年5月30日
開催年月日： 2026年3月10日
記述言語：日本語 会議種別：口頭発表（一般）
-
IMU 統合型 RGB-D センサを用いた豚の体高計測のための 姿勢補正付き 3 次元点群セグメンテーション手法
加藤 一圭、川末 紀功仁、福西 賢大郎、Khin Dagon Win
日本機械学会 九州支部 第79期総会・講演会 2025年5月30日
開催年月日： 2026年3月10日
記述言語：日本語 会議種別：口頭発表（一般）
-
移動式3Dカメラによる黒毛和種肥育牛体重推定の可能性
税所太一，小野芽生，KHIN DAGON WIN，川末紀功仁，徳永忠昭
第18回日本暖地畜産学会佐賀大会 2025年10月26日
開催年月日： 2025年10月25日 - 2025年10月26日
会議種別：口頭発表（一般）
-
バークシャー種肥育豚におけるセンサ情報を用いた体重や産肉形質の経時的変化
小島瑛里香，新美柚樹，浦川啓介，KHIN DAGON WIN，川末紀功仁，徳永忠昭
第18回日本暖地畜産学会佐賀大会 2025年10月26日
開催年月日： 2025年10月25日 - 2025年10月26日
記述言語：日本語 会議種別：口頭発表（一般）
受賞 【 表示 ／ 非表示 】
-
第８回日本オープンイノベーション大賞 「農林水産大臣賞」
2026年2月 内閣府 スカブター：AI・AR 技術による非接触型体重推定デバイスの社会実装
川末 紀功仁、Khin Dagon Win、徳永 忠昭、金子 政時、 助川 慎
-
優秀講演賞
2021年11月 SI2021 機械学習を用いた豚の体重推定
金澤波音、川末紀功仁、Khin Dagon Win
受賞区分：国内学会・会議・シンポジウム等の賞
-
優秀講演賞
2019年11月 SI2019 カメラによる豚の自動体重選別装置
吉田久美子、川末紀功仁、Khin Dagon Win
受賞区分：国内学会・会議・シンポジウム等の賞
科研費（文科省・学振・厚労省）獲得実績 【 表示 ／ 非表示 】
-
オールイン・オールアウト方式養豚におけるペン平均体重の非接触・日次推定基盤の構築
研究課題/領域番号：26K17393 2026年04月 - 2029年12月
科学研究費助成事業 若手研究
Khin Dagon Win
担当区分：研究代表者
-
低出生体重児の非接触身体測定システムの開発
研究課題/領域番号：25K03484 2025年04月 - 2029年03月
科学研究費助成事業 基盤研究(B)
川末 紀功仁、金子 政時、Khin Dagon Win
担当区分：研究代表者
その他競争的資金獲得実績 【 表示 ／ 非表示 】
-
ウェアラブル技術を活用した豚の体重推定システムの革新：精度向上と小型化の挑戦
2025年08月 - 2026年07月
JST Platform for All Regions of Kyushu & Okinawa for Startup-ecosystem(PARKS) Platform for All Regions of Kyushu & Okinawa for Startup-ecosystem(PARKS)
川末 紀功仁、Khin Dagon Win、徳永忠昭、小玉 昂史
担当区分：研究代表者