Image segmentation is a key technology of Digital image processing in the aim of separating the target and the background image classification for the preparation of subsequent identification. Image segmentation as one of the difficulties and the hot in computer vision and image processing, its research by scholars which attached great importance to was the depth and extensive research. Usual image segmentation is always to determine the splitting threshold by subjective,in my paper, it gives a method of fuzzy theory based on the local matrix.This method first constructs the local covariance matrix of the image, and then the application of fuzzy theory, the method is used for image segmentation directly. The one hand, local variance matrix reflects the gray value characteristics and regional changing trends of the smoothing area, texture area and marginal area, the other hand, fuzzy theory is applied to image segmentation techniques, a good solution to the image smoothing area, the texture area and the edge area themselves have a strong fuzzy problem in order to be able to image a more scientific and objective segmentation. |