Research on recognition of motional vehicle based on second-difference algorithm

Location detection method based on vision of motional vehicle is studied in this paper. Conventional method is frame difference method, which is realized by contrasting the different of sequential two frames of vehicle image, and which is easy to detect the background of current frame sheltered by previous frame. Second-difference algorithm is improved in the text, which is the improved algorithm of frame-difference method. Second-difference algorithm locates the motional vehicle in the middle frame relying on the difference among three sequential images. It makes difference operation between two sequential frames and makes difference operation again between two sequential difference images. At last, the results of second- difference are converted to binary for vehicle detection and location. Experimental results show that second-difference algorithm can identify the motional vehicle efficiency and accuracy.