Automated Animal Tracking for Behavioral Experiments
Mingyi Kong, Rong‐Chao Peng · 2022
Animal behavior experiment is essential in neuroscience fields. In order to analyze the trajectory of moving animals using machines instead of manual work in daily experiments, we developed a Matlab algorithm for automatic animal tracking in videos of behavioral experiments. First, the background is extracted using binarization and morphological operations, then the animal is detected and tracked by using the Kalman filter, and finally the motion trajectory and some motion parameters are obtained. We used the algorithm to delineate the motion trajectories of Parkinsonian rats before and during & after cortical stimulation, for evaluation of the efficacy of the cortical stimulation for Parkinson's disease treatment. The results suggest that the developed algorithm is very helpful in analyzing behavioral experiments, as it reduces the heavy manual workload.