Visual attention mechanism based image shadow defect detection

This paper puts forward a new shadow defects segmentation method based on YCbCr color space. First, Mean Shift smoothing method is used to smooth the image pixels, and the motion area which includes the vehicle and the shadow defect is selected by binary discrete wavelet transforms after Mean Shift smoothing, and then the original data of the shadow defect according to the characteristics of the occurrence of shadow is choose, finally, the shape and location of the shadow region is determined by the method which this paper puts forward. The actual road test shows that the method can effectively remove the influence of pedestrians, cyclists in the complex environment, and can obtain integral vehicle shadow defect.