Research On Image Segmentation Based On Improved Fuzzy Clustering

Image segmentation subdivides an image into its sub-areas or objects.The purpose of segmentation is to extract specific targets from complex images.It is an important basis for image recognition,image understanding,and image analysis.With the development of technology,image segmentation has been widely used in many fields,such as medical image processing,face recognition,traffic road analysis,etc.The quality of segmentation directly affects the effectiveness of subsequent work.Therefore,more and more scholars are researching various related image segmentation algorithms.Fuzzy theory can describe the characteristics of image images well,and has been deeply studied in image segmentation.Among them,the most popular segmentation method is the fuzzy segmentation clustering algorithm.After numerous scholars' research and discussion,many improved algorithms have been proposed,but the FCM algorithm itself has some shortcomings,such as sensitivity to the initial clustering center,which is easy to fall into the local The problem of optimality is that the application of spatial position information is insufficient,the noise resistance is not strong,and the number of clusters needs to be manually specified,and the optimization needs to be improved.This article is to optimize and discuss some of the problems of the FCM algorithm itself.This paper studies and analyzes the above problems,and combined with the mainstream intelligent group optimization algorithm,proposes an improved method to improve the efficiency of algorithm segmentation.The main work of this article is as follows:(1)Aiming at the sensitivity of the initial clustering center of the FCM algorithm,a firefly algorithm based on an adaptive step size is used to optimize the fuzzy mean clustering algorithm for image segmentation.The algorithm uses the features of the firefly algorithm to quickly search the global optimum to find The optimal initial clustering center point of FCM.However,due to the fixed step size of the firefly algorithm,the optimal jump problem may occur.Therefore,this article introduces the firefly algorithm and also introduces an adaptive step size control function.Avoid skipping global optimization problems.The adaptive step size can dynamically adjust the step size according to the number of iterations of the algorithm,so that the step size of the algorithm decreases with the increase of the number of iterations.By comparing the optimized algorithm experimentally,the effect of segmentation is better,and The firefly algorithm is analyzed,and the effect of the number of populations on the algorithm is analyzed.Experimental simulations show that the number of populations of the firefly-like algorithm in a certain range has a relatively large effect on the convergence speed of the algorithm.(2)Aiming at the problem that the number of clusters cannot be determined and the stability of the clusters,an improved density peak clustering algorithm is proposed to determine the number of clusters.The algorithm uses the DPC algorithm to automatically determine the number of clusters.The machine selects the clusters.The number of clusters makes clustering more intelligent,but the DPC algorithm does not make full use of the location information of the image,so this paper uses clusters of diamond-shaped neighborhoods to replace the distance of the original pixels.The improved algorithm has a good effect on automatically determining the number of clusters to overcome the stability of artificially determining the number of clusters.