Research On Image Segmentation Algorithm Based On Fuzzy Clustering

Image segmentation is the key link of image processing.The essence of image segmentation is to divide the image into regions.The elements in a region are considered to be similar,and the elements between different regions are quite different.The application of image segmentation is widely used.In medical image processing,the MR image of brain can be segmented,and the distribution of various tissues can be clearly seen,and the abnormal part can be highlighted to help doctors confirm the focus so as to better diagnose and treat the lesions;in the image processing of urban remote sensing image,the layout of the city can be quickly understood and the distribution of the city can be quickly understood.It is of great significance to the construction and development of cities and the monitoring of abnormal range.In addition,the image segmentation can be seen in many other fields.Therefore,in all aspects of life,the quality of image segmentation results are worthy of attention,it will affect the follow-up work frequently,so it is necessary to study the image segmentation based on fuzzy clustering.The image segmentation algorithm used in this paper is based on fuzzy c-means clustering algorithm.Through the analysis of the existing problems of traditional algorithms,the paper considers the methods to solve the problems,and improves the idea and process of the algorithm.Firstly,the paper introduces the technology research of image segmentation and the current situation of FCM algorithm.Then,it focuses on the analysis of the ideas and flow of FCM algorithm and its classical improved algorithm.Then,this paper proposes a FCM algorithm based on particle swarm optimization(PSO?FCM).By introducing PSO algorithm to solve the problem of selecting the initial clustering center and determining the number of clustering categories,and the effectiveness and reliability of PSO algorithm are verified by experiments.At the same time,point out PSO?FCM the limitation of FCM algorithm.In order to solve the problem of noise sensitivity and weak ability of preserving details of traditional algorithms,an improved FCM algorithm based on reliable spatial neighborhood information is proposed in this paper.This algorithm combines spatial neighborhood information,and makes further research on the reliability of neighborhood information,which makes it possible to obtain better results which neighborhood information obtained is more reliable(RSNI?FCM).This improvement greatly improves the anti-noise performance of FCM algorithm,and can obtain higher quality segmentation effect and more accurate segmentation accuracy;at the same time,it retains more details of the image,especially the details of edge pixels,which also makes the segmentation smoother.Finally,through the experimental comparison with FCM algorithm and its derivative algorithm,this paper compares the advantages and disadvantages of RSNI?FCM algorithm and these algorithms,and illustrates the advantages of RSNI?FCM through visual comparison and index parameter comparison.