Unsupervised Texture Segmentation in a Deterministic Annealing Framework

We present a novel optimization framework for unsupervised texture segmentation that relies on statistical tests as a measure of homogeneity. Texture segmentation is formulated as a data clustering problem based on sparse proximity data. Dissimilarities of pairs of textured regions are computed from a multi{scale Gabor lter image representation. We discuss and compare a class of clustering objective functions which is systematically derived from invariance principles. As a general optimization framework we propose deterministic annealing based on a mean{ eld approximation. The canonical way to derive clustering algorithms within this framework as well as an e cient implementation of mean{ eld annealing and the closely related Gibbs sampler are presented. We apply both annealing variants to Brodatz{like micro{texture mixtures and real{word images. T. Hofmann, J. Puzicha, J.M. Buhmann: Unsupervised Texture Segmentation 1