EDGE DETECTION METHOD USING FUZZY MEMBERSHIP FUNCTION

摘要

An object of the present invention is to provide a method capable of effectively detecting edges in a degraded image by detecting edges and performing edge component enhancement as a post-process using a pyramidal region growing technique.;In detecting edges, first segment the image by a pyramid-connected region growth technique, and use the nodes of each level of the pyramid to obtain the subnodes of the upper level of the level. By assigning different weights to each other, the segmentation is performed. After detecting an edge in the segmentation image, the image is transformed into a fuzzy characteristic plane and an edge enhancement function is applied to enhance the edge.;By assigning weights by the membership function to the child nodes used to generate the secondary nodes, the lowpass filtering effect is obtained and a considerable amount of noise can be removed. In addition, by performing edge component enhancement as a post-process, there is an effect that the edge can be effectively detected even in a severely degraded image.