Research on Color Image Edge Detection Based on Neighborhood Change Vector Field
基于邻域彩色变化矢量场的图像边缘检测技术研究*

Firstly, the system framework of image edge detection was discussed and the model of image neighborhood change vector field was established by computing neighborhood changes on multi-direction. Based on the change vector field, this algorithm was used the directional sharpness of neighboring change to describe color edge. And the fuzzy clustering on directional sharpness of pixels in the image was utilized to get the adaptive edge detection result. The results of experiment prove that compared with other edge detection approaches based on gradient, the proposed approach shows good noise robustness and edge location ability.