Visual Detection and Tracking of Pigs : On the Use of Computer Vision in Livestock Farming
With the increasing demand for meat and the corresponding growth in livestock populations, automated monitoring systems are a useful extension of tools to ensure animal welfare. Likewise, such systems enable the processing of larger amounts of data for research and behavioral studies for future welfare recommendations.
This work uses pigs as an example to demonstrate the use of various methods for camera-based detection and tracking of animals. The resulting residence and movement information can be used for targeted study of animal behavior in different experimental housing environments or correlation of movement to other animal health parameters. The results show a high accuracy of the presented methods and thus indicate various practical applications.