The Research Of Digital Image Fuzzy Edge Detection Algorithm

Edges of Image embodies the main profile information of the image, it defines thebrightness, color, and other mutations. Edge detection in image processing technologyis an important direction, widely used in image segmentation, image recognition andimage texture analysis and so on. Images in real life are often influenced by systemnoise, and because of low light, could lead to the result that direct use of image edgedetection operator cannot effectively detect edge, simply smooth filtering processingof the image while at the same time you can remove noise, but it also makes theimage become more blurred, resulting in loss of fuzzy edge detail.On pattern recognition technology,characteristics extraction of image is the key topattern recognition technology, feature extraction carried out on the basis of edgeimage of images is also a research hot, and better results of edge detection for imagesplays an important role in pattern recognition technology. For above problems, thispaper has done the following work：1) this paper has proposed an improved algorithm for image smooth filtering,the filtering algorithm can both eliminate noise and do not damage details informationof fuzzy edge;2) this paper has proposed an improved fuzzy contrast enhanced algorithm，which can make edge differences on both sides more significantly, the algorithm has abetter effect on image fuzzy edge;3) this paper has proposed an improved canny algorithm，The algorithm candetect more weak edges and details.This project finally builds a flexible edge detection platform, the experimentalplatform can achieve fuzzy edge detection for image, and the other can also besatisfactory through interface control of experiment parameters.