Image processing has to deal with much information of an image. Maximum entropy theorem of information theory is one of the useful tools to treat with this kind of information. Based upon the maximum fuzzy partition entropy principle

a novel approach for edge detection is presented. After the concept and the principle of the fuzzy probability and fuzzy partition are introduced briefly

a definition of fuzzy partition entropy is proposed. Using the relation of the probability partition and the fuzzy 2-partition of the image gradient

the algorithm is based on conditional probabilities and fuzzy partition. First

a gradient operator is performed and the gradient image is produced. Second

the problem of edge detection is to find a fuzzy partition of the gradient image

which is considered as being composed of edge region and smooth region

and the automatic optimal threshold is searched from gray-level histogram through maximizing the entropy of fuzzy partition. At last

an edge-enhancing procedure is executed on the edge image. The experiment is conducted on various test images and the results show that the proposed approach has better performances than some classical edge detection methods based on gradient do.