Research Of Co-evolution Particle Swarm Optimization And Its Application In Image Segmentation

Particle swarm optimization (PSO) was presented in 20th century 90s , it's an optimization algorithm based on theory of swarm intelligence. In PSO, swarm intelligence guides the optimization research, which is produced by cooperation and competition among particles in swarm. Compared with evolution algorithm, global searching strategy based on population is reserved, simple speed displacement pattern is adopted and complex genetic operation is avoided in PSO. The memory which PSO has make it track dynamic searching instance to adjust searching strategy. PSO has global convergence ability and robustness and needn't character information of problem. The features of PSO have attracted many scholars, PSO has applied to kinds of fields widely.PSO is based on a hypothesis that information sharing in one swarm, it reflects collaboration among individuals. However, in natural ecosystem, many species improve survival ability through interaction with other species. The interaction exists in all organisms, from cell to advanced biology, exists between animal and plant as well. For conquering the defects of PSO and illumined by co-evolution, co-evolution PSO is proposed.The image segmentation is a key process of the image analysis and the image comprehension. Because of the influence of the complicated background, the object characteristics diversity and the noise, the image segmentation is the difficult and hot research issues on the image processing. The traditional segmentation methods are effective to some images, but are limited to other images which are applied to the especial field and characteristics. Presently, the general segmentation method is not found.Much attention has been paid to the combination of particle swarm optimization and co-evolution, then to create a co-evolution particle swarm optimization algorithm which supporting image segmentation in this paper. The main work is as follows:1. Propose a co-evolution particle swarm optimization algorithmBe illumined by the symbiosis in nature, co-evolution is combined with particle swarm optimization algorithm using Gaussian distribution, create a co-evolution particle swarm optimization algorithm using Gaussian distribution (CPSO-G). In the process of evolution, particles not only exchange the information with other particles in the same swarm, but with particles in other swarms. Gaussian distribution is applied to the generation of random numbers and particle variation which can provides a faster convergence in local search. and increase the diversity of particles so that can help the particle escaping from local extremum. Numeric experiments prove the validity of this algorithm.2.The co-evolution particle swarm optimization algorithm mentioned above is applied to image segmentationThe fuzzy c-means (FCM) clustering algorithm is an effective image segmentation algorithm. But it is sensitive to initial clustering center and membership matrix and likely converges into the local minimum, which causes the quality of image segmentation lower. A new image segmentation algorithm is proposed, which combines the CPSO-G and FCM clustering. Some experiment results are given, which show that the algorithm has the effective ability of searching global optimal solution.