A maximum likelihood estimation method for multispectral autoregressive image models

Author

Bennett, Jesse ; Khotanzad, AIireza

Author_Institution

Dept. of Electr. Eng., Southern Methodist Univ., Dallas, TX, USA

Volume

2

fYear

1997

fDate

26-29 Oct 1997

Firstpage

839

Abstract

We consider the problem of maximum likelihood estimation applied to multispectral random field image models, specifically the multispectral simultaneous autoregressive (MSAR) model. Although previous work has provided least squares methods for parameter estimation, the maximum likelihood method often produces better results. For images with an assumed Gaussian distribution we develop effective procedures for calculating these estimates. Through a series of experiments using known random field models and natural texture samples, the effectiveness of the maximum likelihood approach is demonstrated

Keywords

Gaussian distribution; autoregressive processes; image colour analysis; image sampling; image segmentation; image texture; maximum likelihood estimation; random processes; spectral analysis; Gaussian distribution; color images; experiments; image segmentation; maximum likelihood estimation; multispectral autoregressive image models; multispectral simultaneous autoregressive model; natural texture samples; parameter estimation; random field image models; Color; Gaussian distribution; Image analysis; Image segmentation; Image texture analysis; Lattices; Least squares methods; Maximum likelihood estimation; Multispectral imaging; Parameter estimation; Radio frequency;

fLanguage

English

Publisher

ieee

Conference_Titel

Image Processing, 1997. Proceedings., International Conference on

Conference_Location

Santa Barbara, CA

Print_ISBN

0-8186-8183-7

Type

conf

DOI

10.1109/ICIP.1997.638627

Filename

638627