A TEXTURE-BASED DISTANCE MEASURE FOR CLASSIFICATION

A distance measure based on a new representation scheme of texture ima ges is presented. The new representation scheme captures the structura l and statistical properties of a homogeneous region of texture. Each region is represented by a set of feature frequency matrices (FFM) whi ch gives the frequencies of occurrence of joint feature events. Featur e events are extracted by operators defined by users and/or applicatio ns. The representation is further refined by applying a hierarchical m aximum entropy partitioning scheme to the FFM. The proposed distance m easure is a weighted function of the partitioned FFM. The novelty of t his measure lies in the process of determining the weights. In a class ification experiment, we shall demonstrate the efficacy of the distanc e measure.