Dissertation
Dissertation > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device

Research on Image Segmentation Method Based on Measure of Medium Truth Degree

Author YangZuoZuo
Tutor ZhouNingNing
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
CLC TP391.41
Type Master's thesis
Year 2012
Downloads 14
Quotes 0
Download Dissertation

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 .

Related Dissertations
More Dissertations