Viedo-based Measure of Traffic Volume Parameter

The traffic volume parameters (TVP) measure is the premise to realize the intelligent traffic. In this paper, the video-based measure of TVP is proposed. First, set the detection region which is divided into initiative detection and tracking region. The existence of motion vehicle is tested in the initiative region, if the motion vehicles have appeared, the corners of motion vehicle are located and the state parameters of the motion vehicles are initialized. In the tracing region, the vehicles are traced, according to the kinetic consistency hypothesis of the rigid body, the objections are segmented based on their corners characteristic, in the end of tracing, the information such as the traffic flow and the average velocity etc are computed. The paper also propose corresponding acceleration algorithm of detecting the motion vehicles and locating the corners which speedup the measurement process.