Fast Detection of Vehicles Based-on the Moving Region

摘要

The precise location and tracking of the moving vehicles is an important part in Intelligent Transportation System. This paper presents a new method for detecting vehicles which can remove the shadows fast. The main process is done in three steps: moving region detection, shadow detection, and edge detection. Firstly, a moving region of the image is achieved using the self-adaptive background updating method quickly. Then a coarse shadow area can be taken with the model-based method, and the shadow detection method based on the HSV color space is only applied on the coarse area. Finally, impose edge detection on both the moving regions and shadow areas. Subtraction of the two areas leads to the result of the detection of real vehicle. Experiment results show that this method can improve the efficiency of the moving vehicles detection greatly.