Moving Vehicle Detection Technology Based On Improved Gauss Mixture Model

Moving vehicle detection is an important part of intelligent transportation system. With the maturity and popularization of intelligent transportation system, more and more researchers focus on the domain of moving vehicle detection. For moving vehicle detection in static scenes and the key project of South China University of Technology between Guangdong and Hong Kong “expressway monitor video processing” as background, this paper has done the following work:(1) Based on the comparison of commonly used moving vehicle detection methods, this paper choices Gaussian Mixture Model which has the characteristics of adaptive as the background modeling method, but only use this method cannot get good result when the gray value of moving vehicle and background are very closed, in order to solve this problem, this paper combined with the canny edge detection which is not sensitive to noise and light and can extract complete outline improved the moving object detection result.(2) According to the moving vehicle shadow in some cases, this paper conducts some research on the commonly used shadow processing methods, and compared their features and use cases. The important work is improved the result of elimination algorithm based on YCr Cb color space by combining first order gradient when gray values of moving object and background is very closed. The new method can reduce the misjudgment shadow area.