Wavelet based information for retrieval and classification of mammographic images
Bikesh Kumar Singh, Jitendra Singh Parihar, Pritish Ranjan Pal · 2011
Feature extraction, retrieval and classification of mammographic images are important issues in development of disease diagnostic expert system [DDES]. In this paper we investigate and make comparative analysis of wavelet based CBIR features which may be useful for classification and retrieval of mammographic images for various categories of breast images as classified in Mini-MIAS database. Features were extracted after wavelet decomposition of images rather than from the original image. The results show that extracted features are useful for both retrieval and classification of mammographic images. The precision and classification efficiency may be highly improved by combining extracted statistical features from Approximation, Horizontal, Vertical Detail and Diagonal Detail coefficients.