Gaussian MRF rotation-invariant features for image classification

Author

Deng, Huawu ; Clausi, David A.

Author_Institution

Dept. of Syst. Design Eng., Waterloo Univ., Ont., Canada

Volume

26

Issue

7

fYear

2004

fDate

7/1/2004 12:00:00 AM

Firstpage

951

Lastpage

955

Abstract

Features based on Markov random field (MRF) models are sensitive to texture rotation. This paper develops an anisotropic circular Gaussian MRF (ACGMRF) model for retrieving rotation-invariant texture features. To overcome the singularity problem of the least squares estimate method, an approximate least squares estimate method is designed and implemented. Rotation-invariant features are obtained from the ACGMRF model parameters using the discrete Fourier transform. The ACGMRF model is demonstrated to be a statistical improvement over three published methods. The three methods include a Laplacian pyramid, an isotropic circular GMRF (ICGMRF), and gray level cooccurrence probability features.

Keywords

Markov processes; discrete Fourier transforms; image classification; image texture; Gaussian MRF rotation-invariant features; Laplacian pyramid; Markov random field models; anisotropic circular Gaussian MRF model; discrete Fourier transform; gray level cooccurrence probability features; image classification; rotation-invariant texture features; statistical improvement; texture rotation; Anisotropic magnetoresistance; Discrete Fourier transforms; Feature extraction; Frequency domain analysis; Gabor filters; Image classification; Image texture; Laplace equations; Least squares approximation; Markov random fields; Gaussian MRF (GMRF) model; Markov random field (MRF); anisotropic; classification.; discrete Fourier transform (DFT); isotropic; least squares estimate (LSE); rotational invariance; texture analysis; Algorithms; Artificial Intelligence; Computer Simulation; Image Enhancement; Image Interpretation, Computer-Assisted; Markov Chains; Models, Statistical; Normal Distribution; Pattern Recognition, Automated; Rotation;

fLanguage

English

Journal_Title

Pattern Analysis and Machine Intelligence, IEEE Transactions on

Publisher

ieee

ISSN

0162-8828

Type

jour

DOI

10.1109/TPAMI.2004.30

Filename

1300566