Research And Design Of Medical Image Segmentation Algorithm Based On Fuzzy Clustering

Image segmentation is an important research content in the field of image processing,it is widely used in many fields,such as medicine,meteorology,computer vision,military,remote sensing and so on.Medical image segmentation is a meaningful division of medical images,making it a specified number of different regions,each region is consistent,as different as possible between regions,and the regional mutually disjoint.Medical image segmentation provides useful information for feature extraction and recognition,3D visualization,pathological analysis and diagnosis,etc.The main purpose of medical image segmentation is to divide medical image,so that it have a medical value,and extract a specific area,to facilitate the doctor to develop medical programs,to carry out efficacy evaluation,etc.Medical images in the imaging will be affected by resolution,light conditions,etc.So there is uncertainty,and fuzzy processing technology is just suitable for such problems.This paper deeply studies the basic theory of fuzzy clustering and other artificial intelligence technology,analyzes the problems existing in the previous algorithm and the difficulties encountered in the segmentation of medical images,several improved fuzzy clustering algorithms are proposed and applied to medical image segmentation.In this paper,we obtain the innovative results are as follows:(1)Based on the existing relevant evaluation criteria,this paper presents a new clustering algorithm evaluation criteria: clustering center change value index dv.As a new standard of the performance of clustering algorithm,the dv index is applicable to all clustering algorithms based on clustering center initialization,not just fuzzy clustering,nor is it just image segmentation;attention is what method to find the initial clustering center is better,from the perspective of the results or the overall perspective of the algorithm to assess the advantages and disadvantages of a clustering algorithm.(2)Based on the fast fuzzy clustering algorithm FFCM and En FCM,we propose a more efficient medical image segmentation algorithm.First,we convolute image histogram with Gaussian template,and detect the peak values ofimage histogram,obtain c abscissas of peak values.As a basis for dividing the interval,then initialize and update cluster centers within the scope of each section.Second,mean filter the original image.Last,finalize the segmentation according to interval each pixel value belonging to.Experiments show that,at the premise of ensuring the quality of segmentation,this algorithm can further reduce the running time to more than 4%,better than the current fast segmentation algorithm.(3)Two kinds of new medical image segmentation algorithms based on fuzzy clustering are obtained by using the interval information of image histogram.Firstly,the initial clustering center is used to find the initial clustering center for the initial clustering center or the regional iteration.The method of En FCM is used to update the membership degree,and puts forward the selective calculation of membership degree,which further reduces the computational complexity,improve the efficiency of the algorithm;the range of the corresponding interval is updated by the change of membership degree,and the calculation of the cluster center is restricted in each interval until the change value of the cluster center is smaller than the specified threshold.The iteration ends and the segmentation is completed.Experiments show that the two new algorithms are different from the previous algorithms in different degrees,especially if the second new algorithm is stable in other cases,the efficiency is the best.(4)The qjjc FCM algorithm and gsav FCM algorithm are improved by the selective calculation of membership degree.The former improved better than before the improvement,while the latter improved poor performance.We analyzed the reason,perhaps because of the way to find the initial clustering center and the method of interval partition.