Skip to Main Content
This paper studies some aspects needing to be improved in current video supervision technology. It puts forward video supervision background self-adaptive algorithm in complex environment, by using mathematical morphology, genetic algorithm, rough set theory, etc. We construct morphological structure element according with traffic moving target, and propose mathematical morphology analysis model for traffic video images. At the same time, we study feature extraction based on mathematical morphology and tracking detection methods, and establish typical violation pattern base by utilizing domain expert knowledge.