SEMANTIC CONTENT RECOGNITION AND ORGANIZATION IN MICE BEHAVIORAL TESTING VIDEO

摘要

We present an automatic video analysis system for semantic content recognition in mice testing videos. The system is able to extract from such videos metadata describing the actions and behavior of individual mice. Beginning with the image processing procedures required for identifying the objects of interest, the system tracks the movements and actions of mice in the video sequences, detecting and recording the relevant behavioral events. The metadata obtained, that describe the mice behavior present at the video content, are also organized in a database allowing a by-content accessing of the videos, obtaining important analytical information from the subsequent queries. The database search system enables the retrieving and visualization of selected video sequences matching a given query-by-content criteria. The efficiency and accuracy of the presented system increase analytical power for uncovering and studying the effect of drugs on a variety of behavioral models.