Image Retrieval based on Color Features and Information Entropy
Baohui Zhao, Wenzhun Huang, Haoxiang Wang, Zhe Liu · 2018
In this paper, we conduct research on the novel image retrieval algorithm based on color features and information entropy. Method proposed in this paper with the interception of image center position of the rectangular area use the various levels of color utilization rate, as a feature vector, using color information of color image effectively to reduce the whole picture of the color histogram computation. Using local Euclidean distance finding and similar of sub image library, the key figure in a child in the library each image transform, extract the characteristics of secondary retrieval, output the final retrieval results. Experiments show that using this method can reduce the general amount of calculation of histogram, at the same time can reduce the output image has nothing to do with visual image retrieval, and effectively improves the image retrieval rate, shortening the retrieval time.