Research of image matching based on edge detection and gray level gradient distributing features

摘要

A method based on edge detection and gray level gradient distributing features is presented in this paper. The method aims at solving the problem of objective area ( called ROI) in neighboring images or several frames of images. The objective area can be oriented approximately by the edge detection operator. Then the gray level gradient of pixels distributing on each side of edge line composed of edge pixels can be calculated, and a gray level gradient distributing matrix can be obtained. The objective areas between neighboring images can be matched by the gray level gradient distributing matrixes matching. Experiment results proved the validity of this algorithm.