A real-time vehicle recognition method based on video sequence images

This paper describes a real-time vehicle recognition system in which the basic components of road vehicles are first located in the video sequence images based on background subtraction and then Harris corner of moving vehicles are abstracted. At last, we calculate the Hausdorff distance between the Harris corner of which need to be recognized and that of standard samples of car, bus and truck. The two whose Hausdorff distance is the smallest could be judged as the same type. Because of vehicle detection and recognition in real, cluttered road images, a new vehicle recognition approach is proposed in order to better deal with vehicle variability, illumination conditions, partial occlusions and rotations. The experimental results show that the system can accurately detect and recognize the vehicles on the urban multi-traffic road, while satisfying the real-time requirement.