Multiscale Annealing for Real-Time Unsupervised Texture Segmentation

We derive real{time global optimization methods for several clustering optimization problems commonly used in unsupervised texture segmentation. Speed is achieved by exploiting the image neighborhood relation of features to design a multiscale optimization technique, while accuracy and global optimization properties are gained using annealing techniques. Coarse grained cost functions are derived for central and histogram{based clustering as well as several sparse proximity{based clustering methods. The problem of coarsening sparse random graphs is solved by the concept of structured randomization. For optimization deterministic annealing algorithms are applied. Annealing schedule, coarse{to{ ne optimization and the estimated number of segments are tightly coupled by a statistical convergence criterion derived from computational learning theory. The notion of optimization scale introduced by a computational temperature is thus uni ed with the scales de ned by image and object resolution. The algorithms are benchmarked on Brodatz{ like micro{texture mixtures. Results are presented for an autonomous robotics application. Extensions are discussed in the context of prestructuring large image databases which is necessary for fast and reliable image retrieval. J. Puzicha, J.M. Buhmann: Real{Time Texture Segmentation 1