Research on Human Body Behavior Recognition Based on Vision

Abstract

This paper proposes the research on human body behavior recognition based on vision. Behavior based on high-level human structure can describe behavior more accurately, but it is difficult to extract the behavioral characteristics while often relying on the accuracy of the human pose estimation. Moving object extraction of the moving targets in video analysis as the main content, research based on the image sequence robust, fast moving target extraction, motion estimation and target description algorithm, and the correlation between motion detection is to use frame, frame by comparing the difference between for change and not change area. The model is proposed based on the probability theory, and the future research will be focused on the simulation.