Bivariate shrinkage functions for wavelet-based denoising exploiting interscale dependency

Levent Şendur, Ivan Selesnick

Research output: Contribution to journalArticle

Abstract

Most simple nonlinear thresholding rules for wavelet-based denoising assume that the wavelet coefficients are independent. However, wavelet coefficients of natural images have significant dependencies. In this paper, we will only consider the dependencies between the coefficients and their parents in detail. For this purpose, new non-Gaussian bivariate distributions are proposed, and corresponding nonlinear threshold functions (shrinkage functions) are derived from the models using Bayesian estimation theory. The new shrinkage functions do not assume the independence of wavelet coefficients. We will show three image denoising examples in order to show the performance of these new bivariate shrinkage rules. In the second example, a simple subband-dependent data-driven image denoising system is described and compared with effective data-driven techniques in the literature, namely VisuShrink, SureShrink, BayesShrink, and hidden Markov models. In the third example, the same idea is applied to the dual-tree complex wavelet coefficients.

Original languageEnglish (US)
Pages (from-to)2744-2756
Number of pages13
JournalIEEE Transactions on Signal Processing
Volume50
Issue number11
DOIs
StatePublished - Nov 2002

Fingerprint

Image denoising
Hidden Markov models

Keywords

  • Bivariate shrinkage
  • Image denoising
  • Statistical modeling
  • Wavelet transforms

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Signal Processing

Cite this

Bivariate shrinkage functions for wavelet-based denoising exploiting interscale dependency. / Şendur, Levent; Selesnick, Ivan.

In: IEEE Transactions on Signal Processing, Vol. 50, No. 11, 11.2002, p. 2744-2756.

Research output: Contribution to journalArticle

@article{b04aa2ebe80344b3ac6f97f87d30a083,
title = "Bivariate shrinkage functions for wavelet-based denoising exploiting interscale dependency",
abstract = "Most simple nonlinear thresholding rules for wavelet-based denoising assume that the wavelet coefficients are independent. However, wavelet coefficients of natural images have significant dependencies. In this paper, we will only consider the dependencies between the coefficients and their parents in detail. For this purpose, new non-Gaussian bivariate distributions are proposed, and corresponding nonlinear threshold functions (shrinkage functions) are derived from the models using Bayesian estimation theory. The new shrinkage functions do not assume the independence of wavelet coefficients. We will show three image denoising examples in order to show the performance of these new bivariate shrinkage rules. In the second example, a simple subband-dependent data-driven image denoising system is described and compared with effective data-driven techniques in the literature, namely VisuShrink, SureShrink, BayesShrink, and hidden Markov models. In the third example, the same idea is applied to the dual-tree complex wavelet coefficients.",
keywords = "Bivariate shrinkage, Image denoising, Statistical modeling, Wavelet transforms",
author = "Levent Şendur and Ivan Selesnick",
year = "2002",
month = "11",
doi = "10.1109/TSP.2002.804091",
language = "English (US)",
volume = "50",
pages = "2744--2756",
journal = "IEEE Transactions on Signal Processing",
issn = "1053-587X",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "11",

}

TY - JOUR

T1 - Bivariate shrinkage functions for wavelet-based denoising exploiting interscale dependency

AU - Şendur, Levent

AU - Selesnick, Ivan

PY - 2002/11

Y1 - 2002/11

N2 - Most simple nonlinear thresholding rules for wavelet-based denoising assume that the wavelet coefficients are independent. However, wavelet coefficients of natural images have significant dependencies. In this paper, we will only consider the dependencies between the coefficients and their parents in detail. For this purpose, new non-Gaussian bivariate distributions are proposed, and corresponding nonlinear threshold functions (shrinkage functions) are derived from the models using Bayesian estimation theory. The new shrinkage functions do not assume the independence of wavelet coefficients. We will show three image denoising examples in order to show the performance of these new bivariate shrinkage rules. In the second example, a simple subband-dependent data-driven image denoising system is described and compared with effective data-driven techniques in the literature, namely VisuShrink, SureShrink, BayesShrink, and hidden Markov models. In the third example, the same idea is applied to the dual-tree complex wavelet coefficients.

AB - Most simple nonlinear thresholding rules for wavelet-based denoising assume that the wavelet coefficients are independent. However, wavelet coefficients of natural images have significant dependencies. In this paper, we will only consider the dependencies between the coefficients and their parents in detail. For this purpose, new non-Gaussian bivariate distributions are proposed, and corresponding nonlinear threshold functions (shrinkage functions) are derived from the models using Bayesian estimation theory. The new shrinkage functions do not assume the independence of wavelet coefficients. We will show three image denoising examples in order to show the performance of these new bivariate shrinkage rules. In the second example, a simple subband-dependent data-driven image denoising system is described and compared with effective data-driven techniques in the literature, namely VisuShrink, SureShrink, BayesShrink, and hidden Markov models. In the third example, the same idea is applied to the dual-tree complex wavelet coefficients.

KW - Bivariate shrinkage

KW - Image denoising

KW - Statistical modeling

KW - Wavelet transforms

UR - http://www.scopus.com/inward/record.url?scp=0036844073&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0036844073&partnerID=8YFLogxK

U2 - 10.1109/TSP.2002.804091

DO - 10.1109/TSP.2002.804091

M3 - Article

AN - SCOPUS:0036844073

VL - 50

SP - 2744

EP - 2756

JO - IEEE Transactions on Signal Processing

JF - IEEE Transactions on Signal Processing

SN - 1053-587X

IS - 11

ER -