Pose estimation of sow and piglets during free farrowing using deep learning
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) met...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2024-06-01
|
| Series: | Journal of Agriculture and Food Research |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2666154324001042 |
| _version_ | 1797248302254653440 |
|---|---|
| author | Fahimeh Farahnakian Farshad Farahnakian Stefan Björkman Victor Bloch Matti Pastell Jukka Heikkonen |
| author_facet | Fahimeh Farahnakian Farshad Farahnakian Stefan Björkman Victor Bloch Matti Pastell Jukka Heikkonen |
| author_sort | Fahimeh Farahnakian |
| collection | DOAJ |
| description | 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. |
| first_indexed | 2024-04-24T20:12:26Z |
| format | Article |
| id | doaj.art-cc50e6e8b20841b2b3229c7882d2e44d |
| institution | Directory Open Access Journal |
| issn | 2666-1543 |
| language | English |
| last_indexed | 2024-04-24T20:12:26Z |
| publishDate | 2024-06-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Journal of Agriculture and Food Research |
| spelling | doaj.art-cc50e6e8b20841b2b3229c7882d2e44d2024-03-23T06:25:56ZengElsevierJournal of Agriculture and Food Research2666-15432024-06-0116101067Pose estimation of sow and piglets during free farrowing using deep learningFahimeh Farahnakian0Farshad Farahnakian1Stefan Björkman2Victor Bloch3Matti Pastell4Jukka Heikkonen5Department of Computing, University of Turku, Turku, 20500, Finland; Corresponding author.Department of Computing, University of Turku, Turku, 20500, FinlandDepartment of Production Animal Medicine, University of Helsinki, 00014, FinlandResources Institute Finland (Luke), Latokartanonkaari 9, Helsinki, 00790, FinlandResources Institute Finland (Luke), Latokartanonkaari 9, Helsinki, 00790, FinlandDepartment of Computing, University of Turku, Turku, 20500, FinlandAutomatic 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.http://www.sciencedirect.com/science/article/pii/S2666154324001042Deep learningConvolutional neural networksLivestockPose estimationAnimal behavior |
| spellingShingle | Fahimeh Farahnakian Farshad Farahnakian Stefan Björkman Victor Bloch Matti Pastell Jukka Heikkonen Pose estimation of sow and piglets during free farrowing using deep learning Journal of Agriculture and Food Research Deep learning Convolutional neural networks Livestock Pose estimation Animal behavior |
| title | Pose estimation of sow and piglets during free farrowing using deep learning |
| title_full | Pose estimation of sow and piglets during free farrowing using deep learning |
| title_fullStr | Pose estimation of sow and piglets during free farrowing using deep learning |
| title_full_unstemmed | Pose estimation of sow and piglets during free farrowing using deep learning |
| title_short | Pose estimation of sow and piglets during free farrowing using deep learning |
| title_sort | pose estimation of sow and piglets during free farrowing using deep learning |
| topic | Deep learning Convolutional neural networks Livestock Pose estimation Animal behavior |
| url | http://www.sciencedirect.com/science/article/pii/S2666154324001042 |
| work_keys_str_mv | AT fahimehfarahnakian poseestimationofsowandpigletsduringfreefarrowingusingdeeplearning AT farshadfarahnakian poseestimationofsowandpigletsduringfreefarrowingusingdeeplearning AT stefanbjorkman poseestimationofsowandpigletsduringfreefarrowingusingdeeplearning AT victorbloch poseestimationofsowandpigletsduringfreefarrowingusingdeeplearning AT mattipastell poseestimationofsowandpigletsduringfreefarrowingusingdeeplearning AT jukkaheikkonen poseestimationofsowandpigletsduringfreefarrowingusingdeeplearning |