A novel color feature extraction technique for retrieval of oral pathology images

Abstract

Medical images form an essential source of information for various important tasks such as diagnosis of diseases, surgical planning, medical reference, research and training. A content-based medical image retrieval system has been proposed in this paper which indexes and retrieves images from a digital oral pathology image database. The visual features of the images are extracted based on color content using the novel technique of combining single channel color histogram and Haar wavelet co-efficient. The user supplied query image is compared with that of the images in the database and similar images are retrieved and ranked accordingly. The retrieval of images are compared in CIE L*u*v* color histogram and CIE L-channel histogram. Results show that the single channel color histograms yield better retrieval performance for oral pathology images than the conventional histogram methods.