Research on a novel moving vehicle detection based on video

Video based vehicle detection is an important method of collecting traffic parameter. Its core is moving object detection. In order to overcome the shortcomings of traditional background subtraction based on gray, a background subtraction based on YCbCr color space is proposed to detect moving object and iteration operation to determine the threshold of partition image. The paper proposes grads maximum to find the clearest and the most integrated vehicle image, which provides the basis for vehicle count. Experimental results on real-world throughway sequences show that the proposed method make it possible accurately segment the image, and is very simple and fast. It can be used in relevant real-time applications.