ABSTRACT
Vehicle detection is a popular research in computer vision in recent years. Applying image processing and pattern recognition to traffic management can efficiently reduce traffic accidents, and improve road efficiency, and control pollution. Vehicle detection based on video that has good perspective and extensive usage is the most essential method. For example, it can be used in intelligent transportation systems and driver assistance systems. Video-based vehicle detection is of low cost, easy to install, practical, and has many other advantages, but also stored video is conducive to further human judgment. Because of the diversity as well as the appearance of image noise of the vehicle itself, making vehicle detection becomes a challenging problem. In the light changes, the object occlusion, complex background, vehicle detection research even more complicated.
