Modeling and classifying color textures using random fields in a random environment

We present a random environment model for color textures. This model genera lizes traditional random field models by allowing the spatial interaction p arameters of the field to be random variables. We use this new model to def ine a compact color feature vector which captures within and between color band information. A set of color textures is used to show that this feature vector contains most of the information in the model. We show experimental ly that the new model improves on the performance of multiband Markov field s for texture classification using small samples. (C) 1999 Pattern Recognit ion Society. Published by Elsevier Science Ltd. All rights reserved.