Skip to main content

Computer Vision for animal well-being in the pig sector

This use case facilitates animal well-being assessment in the pig sector by leveraging AgrospAI data sovereignty capabilities. The evaluation is based on video images from pig pens inside the farm. Still, using AgrospAI compute-to-data guarantees farmers that those images are not leaked and that they remain in control.

Computer Vision services derive well-being metrics from the videos, and data consumers just get access to the metrics after the Computer Vision services are executed over the pig pen images in the compute-to-data environment. The metrics include information about how much each pig has moved (or remained static) or how much time they spend in the feeding or drinking areas of the pen.

Image illustrating individual pigs tracking, though pen images are never accessed by consumers, just the computed well-being metrics.

AgrospAI already features a dataset including a sequence of images from video surveillance of one of the pens in the Centre of Swine Studies of Catalonia (CEP), being CEP one of the data space participants:

These images can be used by animal well-being assessment algorithms like the one already featured in AgrospAI. It can be executed on the pig pen images dataset using the compute-to-data provided by Arsys. The algorithm performs automatic image segmentation and tracking to identify and track pig movements in the sequence of images. It also monitors pigs' visits to defined areas of interest, like the automatic feeding machine or the waterer bowl. This allows for the automatic generation of metrics that can be used for animal well-being assessment. The data processing service is available as the