Color feature distribution for content based image retrieval

Features have an important impact on the performance of image retrieval. Wherein spatial features such as shape, color are observatory content, spatial features are observed to more informative. To extract spatial features in image retrieval system, wavelet transformations were used. In conventional content based retrieval system color were taken as a basic feature for representation. However these color features are randomly concentrated on a image sample, and a direct color feature give variance in the color representation. Hence in this paper a color distribution feature using wavelet transformation is proposed. The utilization of wavelet transformation over color feature results in extraction color variation in spectral domain. This finer details give better feature representation of feature, resulting in improved retrieval performance