Presented are video analysis methods for tracking laboratory rats in open field and Y-maze biomedical experiments. The open field test serves as a preliminary assessment in which the rat freely explores a rectangular arena, requiring quantification of visit frequency and duration within designated spatial zones. The next test scenario asses the recognition of novel objects placed on the diagonal of the box by counting visits and their durations for each new object, Finally, the rat movement in a Y-maze environment is analyzed by calculation of visits and visit durations for each branch. The study first introduces enhancements to a previously developed tracking algorithm tailored for the open field scenario. Based on a voting procedure, the method combines results from binarized body detection and identification of the main cluster in the difference image of two consecutive frames. This technique is an alternative to a deep learning-based body localization technique which was the subject of a previous paper. Secondly, a deep learning approach to track the rat in the Y-maze experiment is proposed.