Related Experiment Video
Updated: Aug 8, 2025
06:30
Noninvasive, In-pen Approach Test for Laboratory-housed Pigs
Published on: June 5, 2019
8.5K
Implementation of Computer-Vision-Based Farrowing Prediction in Pens with Temporary Sow Confinement
Maciej Oczak1,2, Kristina Maschat2, Johannes Baumgartner2
1Precision Livestock Farming Hub, The University of Veterinary Medicine Vienna (Vetmeduni Vienna), 1210 Vienna, Austria.
Veterinary Sciences
|February 28, 2023
Summary
Computer vision accurately predicts farrowing by detecting sow nest-building behavior, optimizing temporary confinement for improved animal welfare and piglet survival. This technology reduces labor costs associated with monitoring sows.
Area of Science:
- Animal Science
- Computer Vision
- Agricultural Technology
Background:
- Temporary sow confinement can enhance animal welfare during farrowing.
- Optimal timing for confinement is crucial for sow and piglet well-being.
- Accurate farrowing prediction is needed to implement timely confinement strategies.
Purpose of the Study:
- To predict sow farrowing using computer vision techniques.
- To optimize the timing of temporary sow confinement in farrowing crates.
- To improve animal welfare and piglet survival rates.
Main Methods:
- Utilized the You Only Look Once X (YOLOX-large) computer vision model.
- Detected sow locations and calculated activity levels.
- Applied Kalman filtering and fixed interval smoothing for activity trend analysis.
Main Results:
- Identified the onset of nest-building behavior 12h 51min before farrowing.
- Detected the end of nest-building behavior 2h 38min before farrowing.
- Successfully predicted farrowing for 29 out of 44 sows.
Conclusions:
- Computer vision effectively predicts farrowing by analyzing sow behavior.
- Optimized confinement timing enhances sow and piglet welfare.
- This predictive method can significantly reduce labor costs in swine management.
