Vision-Based Method for Forward Vehicle Detection and Tracking

Vehicle detection and tracking is the basis of advanced driver assistance system. This paper focused on improving the performance of vehicle detection system with single camera and proposed a vision-based method for forward vehicle detection and tracking. The shadow underneath vehicle was segmented accurately by using histogram analysis method and utilized to detect vehicle at daytime. The initial candidates were generated by combining horizontal and vertical edge feature of shadow, and these initial candidates were further verified by using a vehicle classifier based on the histogram of gradient and support vector machine. Kalman filters were used for tracking of the detected vehicles to improve system performance. The results show that the proposed method could be adapt to different illumination circumstances robustly and has a detection rate of 95.78 percent and a false rate of 1.97 percent in normal light condition.