Artificial Intelligence Being Harnessed to Improve Swine Welfare

Researchers with the University of Saskatchewan’s Western College of Veterinary Medicine and Department of Electrical and Computer Engineering are exploring the prospects of using artificial intelligence to improve animal welfare

Research conducted through the office of the NSERC Industrial Research Chair in Swine Welfare is evaluating computer vision technology to score skin lesions on pig carcasses to assess animal welfare.

Dr. Seokbum Ko, a professor in the Division of Biomedical Engineering in the Department of Electrical and Computer Engineering at the University of Saskatchewan, says the goal to develop an automated swine welfare assessment system.

 

 

 “The AI model can identify different areas of the pig and detect lesions on the carcass of specific interest to animal welfare”

 

“The equipment consists of a digital security camera and an artificial intelligence processor. The camera is installed in the abattoir and films the pigs coming down the line following scalding and de-hairing.

“The video from that camera is trained through the AI processor where we  have installed compacted AI models to extract images.

“The model detects the pig carcasses and captures an image of the back of the pig which is dorsal and the side of the pig which is lateral. From this image the AI model can identify different areas of the pig that are of interest, the tail, the flank and the shoulder.

“From these areas of interest, we are now moving forward to train the model to detect lesions on the carcass of specific interest to animal welfare. For this we need to train the model to recognize the different lesions of interest, skin lesions from aggression, a hernia or a tail bite etcetera.

“The end goal is that the models will detect the lesions on the pig carcasses, identify the lesion type and its severity and provide the lesion report for each pig.

“This information would then be collated to provide reports for the abattoir and each barn on the prevalence of lesions in shipments of pigs.“

Dr. Ko says, if we are able to determine a relationship, the automated carcass assessment system will expand animal care assessments by offering information learned from carcasses at slaughter on the handling and management of animals preslaughter and on farm.