Spatially adaptive denoising based on nonstationary correlation assumption

Sung Cheol Park, Moon Gi Kang

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Earlier techniques used for noise filtering were global filters based on the assumption of a stationary image model. Although the stationarity assumption enables the use of the Fast Fourier transform (FFT)-based algorithm, such filters tend to oversmooth edges where stationarity is not satisfied. Recently, various noise smoothing algorithms based on the nonstationary image model have been proposed to overcome this problem. In most conventional nonstationary image models, however, pixels are assumed to be uncorrelated to each other in order not to increase a computational burden too much. As a result, some detailed information is lost in the filtered results. We propose a computationally feasible adaptive noise smoothing algorithm that considers the nonstationary correlation characteristics of images. We assume that an image has a nonstationary mean and can be segmented into subimages that have individually different stationary correlations. Taking advantage of the special structure of the covariance matrix that results from the proposed image model, we derive a computationally efficient FFT-based adaptive linear minimum mean-square-error filter for cases of signal uncorrelated additive, multiplicative, and film-grain noises. The justification for the proposed image model is presented, and the effectiveness of the proposed algorithm is demonstrated experimentally.

Original languageEnglish
Pages (from-to)628-638
Number of pages11
JournalOptical Engineering
Volume43
Issue number3
DOIs
Publication statusPublished - 2004 Mar 1

Fingerprint

Fast Fourier transforms
filters
smoothing
Covariance matrix
Mean square error
Pixels
pixels

All Science Journal Classification (ASJC) codes

  • Atomic and Molecular Physics, and Optics
  • Engineering(all)

Cite this

@article{54932d732dd54cd78f82af869312386c,
title = "Spatially adaptive denoising based on nonstationary correlation assumption",
abstract = "Earlier techniques used for noise filtering were global filters based on the assumption of a stationary image model. Although the stationarity assumption enables the use of the Fast Fourier transform (FFT)-based algorithm, such filters tend to oversmooth edges where stationarity is not satisfied. Recently, various noise smoothing algorithms based on the nonstationary image model have been proposed to overcome this problem. In most conventional nonstationary image models, however, pixels are assumed to be uncorrelated to each other in order not to increase a computational burden too much. As a result, some detailed information is lost in the filtered results. We propose a computationally feasible adaptive noise smoothing algorithm that considers the nonstationary correlation characteristics of images. We assume that an image has a nonstationary mean and can be segmented into subimages that have individually different stationary correlations. Taking advantage of the special structure of the covariance matrix that results from the proposed image model, we derive a computationally efficient FFT-based adaptive linear minimum mean-square-error filter for cases of signal uncorrelated additive, multiplicative, and film-grain noises. The justification for the proposed image model is presented, and the effectiveness of the proposed algorithm is demonstrated experimentally.",
author = "Park, {Sung Cheol} and Kang, {Moon Gi}",
year = "2004",
month = "3",
day = "1",
doi = "10.1117/1.1646181",
language = "English",
volume = "43",
pages = "628--638",
journal = "Optical Engineering",
issn = "0091-3286",
publisher = "SPIE",
number = "3",

}

Spatially adaptive denoising based on nonstationary correlation assumption. / Park, Sung Cheol; Kang, Moon Gi.

In: Optical Engineering, Vol. 43, No. 3, 01.03.2004, p. 628-638.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Spatially adaptive denoising based on nonstationary correlation assumption

AU - Park, Sung Cheol

AU - Kang, Moon Gi

PY - 2004/3/1

Y1 - 2004/3/1

N2 - Earlier techniques used for noise filtering were global filters based on the assumption of a stationary image model. Although the stationarity assumption enables the use of the Fast Fourier transform (FFT)-based algorithm, such filters tend to oversmooth edges where stationarity is not satisfied. Recently, various noise smoothing algorithms based on the nonstationary image model have been proposed to overcome this problem. In most conventional nonstationary image models, however, pixels are assumed to be uncorrelated to each other in order not to increase a computational burden too much. As a result, some detailed information is lost in the filtered results. We propose a computationally feasible adaptive noise smoothing algorithm that considers the nonstationary correlation characteristics of images. We assume that an image has a nonstationary mean and can be segmented into subimages that have individually different stationary correlations. Taking advantage of the special structure of the covariance matrix that results from the proposed image model, we derive a computationally efficient FFT-based adaptive linear minimum mean-square-error filter for cases of signal uncorrelated additive, multiplicative, and film-grain noises. The justification for the proposed image model is presented, and the effectiveness of the proposed algorithm is demonstrated experimentally.

AB - Earlier techniques used for noise filtering were global filters based on the assumption of a stationary image model. Although the stationarity assumption enables the use of the Fast Fourier transform (FFT)-based algorithm, such filters tend to oversmooth edges where stationarity is not satisfied. Recently, various noise smoothing algorithms based on the nonstationary image model have been proposed to overcome this problem. In most conventional nonstationary image models, however, pixels are assumed to be uncorrelated to each other in order not to increase a computational burden too much. As a result, some detailed information is lost in the filtered results. We propose a computationally feasible adaptive noise smoothing algorithm that considers the nonstationary correlation characteristics of images. We assume that an image has a nonstationary mean and can be segmented into subimages that have individually different stationary correlations. Taking advantage of the special structure of the covariance matrix that results from the proposed image model, we derive a computationally efficient FFT-based adaptive linear minimum mean-square-error filter for cases of signal uncorrelated additive, multiplicative, and film-grain noises. The justification for the proposed image model is presented, and the effectiveness of the proposed algorithm is demonstrated experimentally.

UR - http://www.scopus.com/inward/record.url?scp=2342571646&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=2342571646&partnerID=8YFLogxK

U2 - 10.1117/1.1646181

DO - 10.1117/1.1646181

M3 - Article

VL - 43

SP - 628

EP - 638

JO - Optical Engineering

JF - Optical Engineering

SN - 0091-3286

IS - 3

ER -