Survey on Content based Image Retrieval to Deal with Rapid Growth of Digital Images

Abstract

in image retrieval systems has increased in large part due to the rapid growth of the digital images produced by World Wide Web and high capacity of digital data storage devices available in the human race domain. A desired image from network and storage media is shared by citizens belonging to various field, including education, business, government agencies, journalism, and advertising agencies. But due to generation of large collection of digital images, users are not satisfied with the traditional information retrieval techniques. As the elaboration of multimedia technologies are becoming more trendy, so these days the content based image retrieval are becoming a foundation of exact and fast retrieval. This survey paper deals with the techniques of content based image retrieval using both local features and global features. One of them is conventional color histogram and fuzzy color histogram. Further Support vector machine (SVM) with optimized feature sub set selection using radial bias network can also be used to improve retrieval performance.