Research on Fuzzy Clustering Image Segmentation Algorithm based on GA and Gray Histogram

Abstract

When the traditional FCM algorithm is applied to image segmentation, it is sensitive to the selection of clustering center and can not automatically determine the number of clustering center. According to these problems, this paper proposes a fuzzy clustering image segmentation algorithm based on genetic algorithm and gray histogram. This algorithm uses the global search characters of genetic algorithm to avoid falling into local optimal solution, and the objective function is optimized and the number of the cluster center is determined by gray histogram. Experimental results show that the proposed method has strong robustness and good segmentation effect. Introduction Fuzzy c-means clustering algorithm is widely used in the field of image segmentation [1], but it is sensitive to the initial clustering center [4] and can not predict the number of clustering. Genetic algorithm searches the optimal solution by simulating the natural evolution process of the population, and can avoid the local optimal solution. When the gray value is used to calculate the objective function, the repeated calculation is large. Statistical characteristics of the gray histogram can be used to improve the convergence rate, and because the clustering center usually appears in the maximum value of the histogram, the number of clustering center can be determined by the number of maximum values [2]. In this paper, a fuzzy clustering image segmentation algorithm based on GA and gray histogram is proposed. The algorithm uses gray histogram to optimize the objective function and determine the number of clustering center. The global optimal approximate solution of clustering center is obtained by the genetic algorithm [3], and it is used as the initial clustering center of FCM to get the global optimal solution. Image Segmentation based on Traditional FCM The general expression of the objective function is: