Rotation and gray-scale transform invariant texture recognition using hidden Markov model

Author

Chen, Jia-Lin ; Kundu, Amaln

Author_Institution

Dept. of Electr. & Comput. Eng., State Univ. of New York, Amherst, NY, USA

Volume

3

fYear

1992

fDate

23-26 Mar 1992

Firstpage

69

Abstract

In the first stage of the proposed scheme the quadrature mirror filter (QMF) bank is used as the wavelet transform to decompose the texture image into subbands. Gray scale transform invariant features are then extracted from each subband image. In the second stage, the sequence of subbands is modeled as a hidden Markov model (HMM), and one HMM is designed for each class of textures. During recognition, the unknown texture is matched against all models. The best matched model identifies the texture class. The angles of rotation in the experiments are selected randomly in between -90° and 90°. Up to 95% classification accuracy is reported

Keywords

digital filters; filtering and prediction theory; hidden Markov models; image texture; wavelet transforms; HMM; QMF bank; gray-scale transform invariant texture recognition; hidden Markov model; quadrature mirror filter; rotation-invariant texture recognition; subbands; texture image; wavelet transform; Bandwidth; Feature extraction; Filter bank; Finite impulse response filter; Frequency; Gray-scale; Hidden Markov models; Humans; Visual system; Wavelet transforms;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech, and Signal Processing, 1992. ICASSP-92., 1992 IEEE International Conference on

Conference_Location

San Francisco, CA

ISSN

1520-6149

Print_ISBN

0-7803-0532-9

Type

conf

DOI

10.1109/ICASSP.1992.226274

Filename

226274