Research on Image Segmentation Optimization Algorithm based on Chaotic Particle Swarm Optimization and Fuzzy Clustering

Abstract

Fuzzy C-Means clustering, FCM, is an unsupervised learning algorithm. The algorithm is easily affected by noise points and depends on the initial values. When the sample value is large, the algorithm is easy to fall into local extremum. In this study, the traditional fuzzy clustering algorithm is improved, and the particle swarm optimization algorithm with global optimization ability is applied to the FCM algorithm, and chaotic technology is added. Chaotic variables produce a chaotic sequence based on the current global optimal position, using chaotic sequence has the best fitness value of particles randomly instead of a particle of the particle swarm, the improved algorithm can effectively avoid the stagnation of particles in the iteration, fast search to the global optimal solution, avoid convergence to local extremum. Experimental results indicate that this algorithm overcomes the dependence on the initial clustering centre of FCM, which brings high robustness and segmentation accuracy, and has more faster convergence speed.