Automatic tracking analysis in morris water maze biomedical vidoes

We present an automatic video processing system for tracking analysis in the Morris water escape testing videos. This system is able to extract automatically from such videos information (metadata) describing the spatio-temporal trajectories of animals in the maze and the timings of behavioral events such as stopping or crossing of a target area. The specific semantic metadata are produced by the system and saved in a XML file that describes the mice behaviour extracted from the video content. This description is also stored in a database using a data model that allows one to perform subsequent queries to obtain important factual and analytical information and to retrieve and visualize selected video sequences matching specific query criteria. In all cases, we have observed a significant increase in both accuracy and efficiency when undertaking such analysis using the procedures described in this paper, in comparison with current methods