Research On Image Segmentation Method Based On Measure Of Medium Truth Degree

Image segmentation is just to divide an image into different sub-images with different characters and extract some interested objects. It is a key step from image processing to image analysis, and is also a classical difficult problem.This thesis studies the fuzzy clustering theory, the standard FCM algorithm and the FCM image segmentation algorithm and analyzes the problems of the FCM image segmentation algorithm. The FCM image segmentation algorithm is sensitive to noise because of not taking into account the spatial information of the image. It uses the degree of gray similarity and correlation of the pixel and its neighbors and the measure of medium truth degree to define the medium membership function, and proposes a FCM image segmentation algorithm based on measure of medium truth degree. Experimental results show that the proposed algorithm is more robust than the standard FCM image segmentation algorithm, with more certainty and less fuzziness.Based on the deep study of the classical edge detection algorithms, this work uses two features of images as follows: the gray value of the edge point changes significantly in its neighborhood, and there is a local peak in vertical direction of the image edge. Considering an image edge detection algorithm based on measure of medium truth degree and taking into account of the iterative threshold method, it proposes an adaptive image edge detection algorithm based on measure of medium truth degree. Experimental results show that the new algorithm can not only detect the continuous edge, but also can successfully preserve the image edge details and is more robust compared with that of the classical algorithms.