Abstract: Video surveillance system is a component of intelligent transportation system.Through preprocessing and detecting the video image of vehicles in monitoring area,and real-time classification,the release time of traffic lights can be confirmed in real time according to the size of the vehicle flow.By using the method of the active contour tracking model and fuzzy support vector machine,the intelligent traffic video surveillance system provides real-time detection and classification of vehicles.Matlab simulation results show that the average correct recognition rate of large,medium and small sized vehicle is  8 9.8%,so the method can improve the efficiency of traffic.

Key words: difference image, active contour tracking model, fuzzy, support vector machine