ETRI-Knowledge Sharing Plaform

ENGLISH

성과물

논문 검색
구분 SCI
연도 ~ 키워드

상세정보

학술지 Estimation of Number of Pigs Taking in Feed Using Posture Filtration
Cited 3 time in scopus Download 101 time Share share facebook twitter linkedin kakaostory
저자
김태호, 김유진, 김세한, 고재필
발행일
202301
출처
Sensors, v.23 no.1, pp.1-14
ISSN
1424-8220
출판사
MDPI
DOI
https://dx.doi.org/10.3390/s23010238
협약과제
22HR7100, 축산질병 예방 및 통제 관리를 위한 ICT 기반의 지능형 스마트 안전 축사 기술 개발, 김세한
초록
Pork production is hugely impacted by the health and breeding of pigs. Analyzing the eating pattern of pigs helps in optimizing the supply chain management with a healthy breeding environment. Monitoring the feed intake of pigs in a barn provides information about their eating habits, behavioral patterns, and surrounding environment, which can be used for further analysis to monitor growth in pigs and eventually contribute to the quality and quantity of meat production. In this paper, we present a novel method to estimate the number of pigs taking in feed by considering the pig's posture. In order to solve problems arising from using the pig's posture, we propose an algorithm to match the pig's head and the corresponding pig's body using the major-and-minor axis of the pig detection box. In our experiment, we present the detection performance of the YOLOv5 model according to the anchor box, and then we demonstrate that the proposed method outperforms previous methods. We therefore measure the number of pigs taking in feed over a period of 24 h and the number of times pigs consume feed in a day over a period of 30 days, and observe the pig's feed intake pattern.
KSP 제안 키워드
Anchor box, Breeding environment, Eating habits, Feed intake, Meat production, Minor axis, Quality and quantity, Supply Chain Management(SCM), Surrounding environment, behavioral patterns, detection performance
본 저작물은 크리에이티브 커먼즈 저작자 표시 (CC BY) 조건에 따라 이용할 수 있습니다.
저작자 표시 (CC BY)