Research On Image Segmentation Method Based On Improved PSO-FCM Clustering

Image segmentation is an important step in image processing.The quality of the segmentation results has a great influence on the subsequent image processing.At present,image segmentation is widely used in many fields such as industrial and agricultural production,manufacturing,transportation planning and medical diagnosis.Image segmentation based on fuzzy clustering has attracted wide attention from scholars because it can better deal with the fuzzy information in the image,does not require too much human interference and is easy to implement.However,the traditional fuzzy C-means clustering(FCM)also has defects,such as sensitive to the initial clustering center and easy to fall into local optimum.In order to give full play to the advantages of FCM clustering algorithm in image segmentation,this dissertation studies and improves the image segmentation method based on FCM clustering algorithm.Firstly,in order to improve the problem that the standard particle swarm optimization(PSO)is easy to fall into local optimum prematurely and its convergence speed is slow in the later stage,an improved PSO algorithm is proposed in this paper.It uses Halton sequence to initialize the population,and introduces an improved inertia weight based on chaotic mapping and a nonlinear dynamic learning factor.Then this paper verifies its feasibility and effectiveness through simple experiments.Secondly,this paper combines the improved PSO algorithm with FCM clustering to form an improved PSO-FCM algorithm.The objective function of FCM is used as the fitness function of PSO algorithm.The clustering center of FCM algorithm is updated by the position update formula of PSO algorithm.The optimal position with the smallest fitness value is the optimal clustering center.Finally,the improved algorithm is applied to gray image segmentation.By comparing with the standard FCM and PSO-FCM segmentation results,and comparing the PSNR,SSIM and FSIM of the segmentation results,it is proved that the image segmentation algorithm has higher segmentation accuracy and convergence speed.