Image Retrieval Combining FOA Optimized PCNN and Phase Congruency

For resolving the problem that area and edge information can not be fused efficiently in content-based image retrieval, an image retrieval method combining optimized pulse coupled neural network（PCNN） and phase congruency（PC） is presented. Firstly, three key parameters of simplified PCNN model are set adopting fruit fly optimization algorithm（FOA）, and image segmentation and feature extraction are performed based on the PCNN model and maximum information entropy criterion. Then image edge is extracted utilizing PC, and edge color histogram feature is obtained. Finally, image retrieval is achieved by integrating the two types of image features. Experimental results demonstrate that the proposed method has good retrieval performance, and its average precision and recall are higher than representative methods.