Automating monitoring of cat feeding behaviour

Abstract

Cat food manufacturers spend a significant proportion of their research budget on food formulation and palatability. In this paper we propose an efficient and economic method of monitoring cat feeding behaviour during palatability trials. Instrumenting food bowls with load cells measures how much is eaten in each meal, and video records the interaction with the food during meals. Adaptive background subtraction is used to trigger recording, eliminating long periods of animal inactivity and uninteresting video. A single computer is able to monitor and record 4 cages simultaneously at 25 frames per second.