Digital imaging for animal monitoring progresses

Camera and vision techniques to identify and monitor individual animals in a group still need improvement.

Earlier in July, 60 people participated in a Breed4Food seminar in Wageningen, Netherlands, on the use of sensors and cameras for recording traits and monitoring animals in groups.

According to an announcement from Wageningen University & Research (WUR), participating researchers shared their results and experiences, after which industry representatives gave their viewpoints on implementation issues.

The seminar demonstrated the added value of cooperation between research and practical implementation for sustainable animal production, WUR said.

Individual animal challenges

Animals are housed in groups, but for breeding, it is best to measure animal performance individually, WUR said, adding that it would be ideal to collect longitudinal data by recording the performance of animals in their daily routine, without human interference.

The first session showed the results of different techniques to record: (1) the activity of broilers, which does depend on weight class and age, and (2) the steps, locomotion and weight of turkeys using different sensors such as IMU tags, video imaging and force plates.

WUR said the first results are promising for video imaging, supporting the concept that digital tools can do what a human eye can see (and can do that even better).

The research also showed that it is not yet straightforward to use cameras and vision techniques to detect, identify and monitor individual animals in a group, WUR said. Follow-up work is still needed, as are running projects like GenTORE (H2020), SmartResilience (NWO) or possible new projects in the (near) future, WUR said.