Research on Image Retrieval Based on Fusion the Scale Free Particle Swarm Optimization

This paper propose a Tabu search particle swarm algorithm which fusions the Scale Free Like network to optimize the improved image retrieval relevance feedback.This method opitimizes the feedback process by the Tabu search particle swarm algorithm which fusions the Scale Free Like network and lead into resolution ratio through user's feedback information to adjust the feature weight and the query vector dynamically.Experiment shows that the method can not only increase the diversity of particle population, avoid premature convergence, enhance the ability of particle to jump out of local optimal solution, but also retrieve pictures that more user-friendly, effectively improved the retrieval efficiency and retrieval accuracy.