Research on Image Segmentation Method Based on Measure of Medium Truth Degree
|School||Nanjing University of Posts and Telecommunications|
|Course||Applied Computer Technology|
|Keywords||Image Segmentation FCM image segmentation algorithm Intermediary measure of the extent of the true value Intermediary membership function Edge Detection|
Image segmentation is the image into each with characteristics of the area and extracted interested in the technology and process . It is a critical step of the image processing to the image analysis , is a classic puzzle . In this paper, the theory of fuzzy clustering , the standard FCM algorithm and FCM algorithm based image segmentation method conducted in-depth research and FCM image segmentation algorithm difficulties . FCM image segmentation algorithm sensitive to noise problems , consider the space between the image pixels , pixel and its neighbor similarity and relevance and degree of intermediary true value of the measure of the domain pixel gray the intermediary membership function is defined , FCM image segmentation algorithm based on the degree of intermediary true value to measure the . Experimental results show that the noise immunity of the algorithm is significantly better than the standard FCM image segmentation algorithm , and with greater certainty and vagueness in the time of the split . The classic image edge detection algorithm -depth study on the basis of two characteristics of the image edge : edge of the neighborhood of the point of gray value changes significantly toward a vertical direction and the edge with the edge of a local peak , combined iterative calculation method of image edge detection algorithm and threshold metrics based on the degree of intermediary true value , designed to measure adaptive image edge detection algorithm based on the degree of intermediary true value . Experimental results show that the new algorithm is not only capable of detecting a continuous edge , and more effective than the classic edge detection algorithm preserving image details and better noise immunity .