Color Coarse Segmentation and Regions Selection For Similar Images Retrieval

With the growth of large image databases, content-based image retrieval systems are actually a highly challenging problem. The common approach is to extract a signature for every image based on different features (texture, color, shape analysis ) and to minimize a distance for retrieving similar images to a request one. Then, features extraction becomes the most important theme objectively , a large panel of systems [4] [12] and methods exist, based on statistical features[3], visual parameters, color histograms[10], region-based search[8] The main attention must be paid to develop insensitive features to intensity variation, scaling, rotations or else compression effects. Finally, we will develop the solution1 to extract some numerical features for every image before achieving with the presentation of our content-based retrieval system called iCOBRA2. The efficiency of this method will be illustrated on a large classical color images database, composed notably by goodshoot©images, containing very diverses images with a high rate of jpeg compression.