Application of Improved Fuzzy Clustering Method in the Image Segmentation

This paper presents an operator of fuzzy clustering method of image segmentation based on Local Binary Pattern (LBP). Semi-supervised learning and fuzzy clustering method are introduced in order to overcome the problem of initial clustering sensitive. Also, local binary pattern operator is introduced to construct the space feature vectors of pixels, which makes full use of the space characteristics information of pixel. Space feature vectors of pixels are used to choose image segmentation clustering data set. Result of experiments shows that the method improves the accuracy of segmentation and enhance the detailed features of segmentation.