Extraction of moving object in video sequences of a dynamic scene based on gray

An improved moving object segmentation method is presented based on gray in the video sequence. The Sobel is improved and the edge is extracted. The image is divided up into several connected areas based on the edge. The feature points are selected in the connected areas and are matched for connected areas merging and the global motion parameter. Motion region is obtained by using the frame difference intersection of adaptive frame interval in corrected image sequence. The motion object is extracted based on the relation between the connected area and motion region utilizing. The experimental results show that the algorithm is reasonable and effective in moving objective segmentation in video sequence in the dynamic background.