Abstract: In the last decades has witnessed the research interest in the field of Content Based Image Retrieval (CBIR). It inspects the visual content and detects images in the database. In this study, it is proposed a novel approach to improve the efficiency of the CBIR system based on heuristic Particle Swarm Optimization (PSO). The mean and standard deviation values of the color images are used as color features. Moreover, it is also contemplated an active contour as the region feature. PSO is used to reduce the semantic gap between retrieval results and user expectations. Experiments demonstrated that the proposed image retrieval system performed well.