IJSRD - International Journal for Scientific Research & Development| Vol. 3, Issue 07, 2015 | ISSN (online): 2321-0613

Content based image retrieval is one of the extremely standard and growing research areas of the DIP (Digital Image Processing). Most of the offered image search tools, for instance Google Images and Yahoo! Image search, are centered on textual annotation of digital images. In these search techniques/tools, digital images are manually annotated with keywords and then recovered with the help of text-based search approaches. The performances of these schemes are not acceptable. In this paper, we will suggest a unique and efficient system for the image retrieval which is based on extracting color feature of an image. Suggested system utilizes wavelet transform and K means algorithm for image decomposition & feature extraction, it also uses Euclidian similarity function for the feature comparison. It is proved by the experiment outcomes the efficiency of the suggested image retrieval system is enhanced in comparison with the existing image retrieval systems.