A fast detection algorithm for moving vehicle in traffic scenes

The vehicle detection is extremely important for the traffic parameter extraction and analysis in traffic scenes. In order to detect moving vehicle quickly, this paper proposes a samples-based adaptive segmentation detection algorithm. First, select the first frame as a reference frame, use a random policy to select values to build a samples-based estimation of the background. Second, we adopt different strategies to update different regions, and use frames subtraction and background extend detection classify the pixels in each frame into background area, uncovered background area and background extend area. Then we use different rates to update these areas. The experimental results indicate that our algorithm achieves good performance in real-time properties, quickly and accurately segment the moving vehicles.