A multiresolution hybrid neuro-Markovian image modeling and segmentation

Abstract

29 285 Brest Cedex w a r s a ABSTRACTA new textured image model is proposed. This model is described by means of a neuro-markovian hybrid approach using a Kohonen map and a Hidden Markov Model (HMM). Each state of the HMM describes one resolution level in the image. The change of the state corresponds to the change of image analysis resolution level. The HMM observation space is composed of clusters which are estimated using a Kohonen map. The second role of the Kohonen algorithm is to achieve a segmentation which is done in parallel with the one proceeded by a Viterbi algorithm. The results given by the both algorithms present some complementarity.