An Edge Detection Algorithm Based on Adaptive Weight Multi-structure Element Mathematical Morphology

摘要

The key of image edge detection is to detect the ed of the image as much as possible meanwhile restrain noi effectively. Aimed to achieve this objective, this paper h proposed adaptive weight multi-structure eleme mathematical morphology edge detection algorithm based i local images. This method has solved the image edge detecti distortion, which is resulted from the whole structures weig acting on the local images. The main steps are as the followin segment the image, binary local image via OSTU, then dete the edge using adaptive multi-structure mathematic morphology. The experiment results demonstrated that tt method reflects the edge direction of original ima information accurately, showing better noise immunity ai accuracy.