An adaptive edge detection algorithm based on gray-scale morphology

Abstract

The method of image edge detection algorithm based on the traditional mathematical morphology has good ability to filter noise. But the method of using single form and single scale form element to edge the image has pertinence so the generality is weak. An algorithm of adaptive grayscale image edge detection based on multi-scale and multi-form is proposed in this paper. The image is processed with different shapes and dimensions of structural elements, and the adaptively weights are calculated according to information entropy. And we build a series-parallel form of edge detector. At last, we can get the final edge image after the fusion of each result from single element according to the series-parallel edge detector. The results of simulation show that compared with several conventional edge detection algorithms, the proposed algorithm can not only filter most of noise but also retain good edge details.