A multiresolution hybrid neuro-Markovian image modeling and segmentation

Author

Wiliñski, Piotr ; Solaiman, Basel ; Hillion, Alain ; Czarnecki, Witold

Author_Institution

Ecole Nat. Superieure des Telecommun. de Bretagne, Brest, France

Volume

3

fYear

1996

fDate

16-19 Sep 1996

Firstpage

951

Abstract

A 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

Keywords

hidden Markov models; image classification; image resolution; image segmentation; image texture; parameter estimation; self-organising feature maps; HMM observation space; Kohonen map; clusters; contextual classification; hidden Markov model; image segmentation; multiresolution image analysis; neuro-Markovian hybrid approach; textured image model; Clustering algorithms; Hidden Markov models; Image analysis; Image resolution; Image segmentation; Image texture analysis; Information analysis; Pixel; Signal processing algorithms; Viterbi algorithm;

fLanguage

English

Publisher

ieee

Conference_Titel

Image Processing, 1996. Proceedings., International Conference on

Conference_Location

Lausanne

Print_ISBN

0-7803-3259-8

Type

conf

DOI

10.1109/ICIP.1996.560957

Filename

560957