Research of image edge detection and application based on grey prediction model

On the basis of grey prediction theory, a new method of image edge detection is put forward. First, the method of edge detection based on grey GM(1,1) prediction model and its realizing principles of this method are discussed, and the image edges can be detected from eight directions by means of this method. Moreover, the corresponding simulation experiment is implemented and its ability of edge detection is tested. At last, this method is used to detect the edge of licence plate. Compared with other methods, the result shows that this algorithm can make the character segmented from the background distinctly, and advance accuracy of vehicle license plate recognition effectively.