Image denoising using the non-uniform directional filter bank and hidden markov tree

Author

Bui, Ha H. ; Nguyen, Truong T.

Author_Institution

Electr. Eng. Dept., Polytech. Univ., Brooklyn, NY

fYear

2006

fDate

24-27 Sept. 2006

Firstpage

214

Lastpage

217

Abstract

The non-uniform directional filter bank (nuDFB) is a new two-dimensional decomposition of image that have both multiresolution and multidirectional properties. In this paper we proposed a hybrid filter bank that combine the nuDFB with the discrete wavelet transform (DWT). The new image decomposition is then used in image denoising application, where the hidden Markov model is applied to account for the inter-scale, inter-direction, and inter-location dependency

Keywords

channel bank filters; discrete wavelet transforms; hidden Markov models; image denoising; image resolution; trees (mathematics); DWT; discrete wavelet transform; hidden Markov tree; hybrid filter bank; image denoising; multidirectional properties; multiresolution properties; nonuniform directional filter bank; nuDFB; two-dimensional image decomposition; Discrete wavelet transforms; Filter bank; Frequency; Hidden Markov models; Image decomposition; Image denoising; Image reconstruction; Image resolution; Signal resolution; Wavelet coefficients;

fLanguage

English

Publisher

ieee

Conference_Titel

Digital Signal Processing Workshop, 12th - Signal Processing Education Workshop, 4th

Conference_Location

Teton National Park, WY

Print_ISBN

1-4244-3534-3

Electronic_ISBN

1-4244-0535-1

Type

conf

DOI

10.1109/DSPWS.2006.265409

Filename

4041061