Modified Fuzzy C-Means Clustering Algorithm Of Image Segmentation

Image segmentation is dividing an image into a number of specific areas with unique properties according to certain criteria and extract the part of people's interest.Image segmentation is an important part of image processing and application.It is an important basis for image recognition,understanding and analysis.With the rapid development of modern information technology,image segmentation technology plays an increasingly important role in social development.It has also been increasingly integrated into many aspects of social life,traffic analysis,scene-specific analysis and medical image analysis.The quality of image segmentation directly affects the effectiveness of follow-up work.Therefore,image segmentation is an important technology.Researchers have done a lot of research on image segmentation,but so far there is no universal segmentation method.Image segmentation methods can be divided into area-based image segmentation,edge detection based image segmentation,graph theory based image segmentation and clustering based image segmentation.Fuzzy C-means clustering algorithm(FCM)is a very important and widely used method for image segmentation in clustering.The FCM algorithm is mainly aimed at fuzzy and uncertain images.When the image is segmented,the problem of threshold setting is avoided.No manual intervention is needed in the clustering process,and automatic image segmentation is implemented.The classic FCM algorithm for image segmentation still has some problems.Therefore,for the determination of clustering number and cluster center in FCM algorithm,the sensitivity to initial value is easy to fall into local extremum,the use of spatial information,the poor anti-noise ability and the sensitivity to linear indistinguishable data are improve.In this paper,we first use the differential evolution two-dimensional entropy algorithm to initialize the image to get the center of the target and the background,as the initial clustering center of FCM algorithm,image segmentation by KFCM algorithm.This method improves the segmentation speed of the FCM algorithm,because of the limited selection of clustering number,not suitable for medical image segmentation.To solve this problem,an improved non-local extreme value fuzzy C-means clustering algorithm is proposed in this paper.The proposed algorithm finds the clustering center and the number of clusters by calculating the slope of each point of the histogram,according to its rules to determine the cluster center and cluster number,to solves the problem that the initial value is sensitive and easy to fall into the local optimal solution.Secondly,the non-local filter is introduced to compute the weighted image,the gray information and spatial information are combined to suppress the noise of each pixel in the non-local spatial information image and improve the segmentation accuracy.Finally,according to the principle of the maximum degree of membership,the image pixels areclassified to complete efficient segmentation.Through the experimental verification of medical images,calculate JS index to quantitatively analyze the segmentation accuracy.The experimental result shows that the algorithm can effectively remove the noise and keep the details of the image well,enhance the robustness of the segmentation and improve the segmentation precision.