COMBINING IMAGE COMPRESSION AND CLASSIFICATION USING VECTOR QUANTIZATION

Statistical clustering methods have long been used for a variety of signal processing applications,including both classification and vector quantization for signal compression. We describe a method of combining classification and compression into a single vector quantizer by incorporating a Bayes risk term into the distortion measure used in the quantizer design algorithm. Once trained, the quantizer can operate to minimize the Bayes risk weighted distortion measure if there is a model providing the required posterior probabilities, or it can operate in a suboptimal fashion by minimizing only squared error. Comparisons are made with other vector quantizer based classifiers, including the independent design of quantization and minimum Bayes risk classification and Kohonen's LVQ, A variety of examples demonstrate that the proposed method can provide classification ability close to or superior to LVQ while simultaneously providing superior compression performance.