Multiresolution GMRF models for texture segmentation

Author

Krishnamachari, Santhana ; Chellappa, Rama

Author_Institution

Dept. of Electr. Eng., Maryland Univ., College Park, MD, USA

Volume

4

fYear

1995

fDate

9-12 May 1995

Firstpage

2407

Abstract

A multiresolution model for Gauss Markov random fields (GMRF) is presented. Coarser resolution sample fields are obtained by either subsampling or local averaging the sample field at the fine resolution. Although Markovianity is lost under such resolution transformation, coarser resolution non-Markov random fields can be effectively approximated by Markov fields. We use a local conditional distribution invariance approximation, to estimate the parameters of the coarser resolution processes from the fine resolution parameters. This multiresolution model is used to perform texture segmentation

Keywords

Gaussian processes; Markov processes; image resolution; image sampling; image segmentation; image texture; invariance; parameter estimation; GMRF models; Gauss Markov random fields; coarser resolution sample fields; fine resolution parameters; image segmentation; local averaging; local conditional distribution invariance approximation; multiresolution model; parameter estimation; subsampling; texture segmentation; Automation; Covariance matrix; Educational institutions; Gaussian processes; Image segmentation; Lattices; Markov random fields; Parameter estimation; Probability density function; Spatial resolution;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech, and Signal Processing, 1995. ICASSP-95., 1995 International Conference on

Conference_Location

Detroit, MI

ISSN

1520-6149

Print_ISBN

0-7803-2431-5

Type

conf

DOI

10.1109/ICASSP.1995.479978

Filename

479978