Pose estimation of sow and piglets during free farrowing using deep learning

被引：5

作者：

Farahnakian, Fahimeh ^{[1
]}

Farahnakian, Farshad ^{[1
]}

Bjorkman, Stefan ^{[2
]}

Bloch, Victor ^{[3
]}

Pastell, Matti ^{[3
]}

Heikkonen, Jukka ^{[1
]}

机构：

[1] Univ Turku, Dept Comp, Turku 20500, Finland

[2] Univ Helsinki, Dept Prod Anim Med, Helsinki 00014, Finland

[3] Resources Inst Finland Luke, Latokartanonkaari 9, Helsinki 00790, Finland

来源：

JOURNAL OF AGRICULTURE AND FOOD RESEARCH | 2024年 / 16卷

关键词：

Deep learning; Convolutional neural networks; Livestock; Pose estimation; Animal behavior; AUTOMATIC RECOGNITION; BEHAVIOR; CLASSIFICATION; MANAGEMENT; SYSTEM; PEN;

D O I：

10.1016/j.jafr.2024.101067

中图分类号：

S [农业科学];

学科分类号：

09 ;

摘要：

Automatic and real -time pose estimation is important in monitoring animal behavior, health, and welfare. In this paper, we utilized pose estimation for monitoring the farrowing process to prevent piglet mortality and preserve the health and welfare of the sow. State -of -the -art Deep Learning (DL) methods have lately been used for animal pose estimation. This paper aims to probe the generalization ability of five common DL networks (ResNet50, ResNet101, MobileNet, EfficientNet, and DLCRNet) for sow and piglet pose estimation. These architectures predict the body parts of several piglets and the sow directly from input video sequences. Real farrowing data from a commercial farm was used for training and validation of the proposed networks. The experimental results demonstrated that MobileNet was able to detect seven body parts of the sow with a median test error of 0.61 pixels.

引用

页数：13

共 60 条

[1] Artificial intelligence in animal farming: A systematic literature review [J].