HISTOGRAM APPROACH FOR CONTENT-BASED IMAGE RETRIEVAL

摘要

Color histograms are widely used for the Content-Based Image Retrieval (CBIR). In this paper we propose a new image retrieval technique that combines color and edge features for image indexing. We employ the perceptually uniform HSV (Hue, Saturation and Value) color space for color histograms and YCrCb (luminance/red chrominance/blue chrominance) color space for edge histograms in our work. We use Euclidean distance for distance measurement between a query image and images in a database and we have also explained the way to switch between different set of features, different color spaces and different distance measurements. We have used Matlab environment to implement this system. Our approach toward this system has been more experimental based than hypothetical.