A major design issue in content-based image retrieval systemis the selection of the feature set. This study attacks the problemof finding a discriminative feature for each class, which is optimalin some sense. Fuzzy ARTMAP architecture is used to find thisdiscriminative feature set. For this purpose, initially, a largevariety of features are extracted from the regions of thepre-segmented images. Then, the feature set of each object class islearned using the Fuzzy Art Map Architecture, by identifying theweights of each feature for each object class. In the querying phase,trained set of feature weights of fuzzy ARTMAP's are used to find thelabel of each object class. This task is achieved by combining theregions in the images and computing the maximum membership value forthe compound regions, which correspond to a possible object class. The query object is matched to each segment group in a fuzzy databaseusing the membership values of segment groups.
展开▼
