Dual learning based compression noise reduction in the texture domain

Jae Won Lee, Oh Young Lee, Jong-Ok Kim

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Compression noise reduction is similar to the super-resolution problem in terms of the restoration of lost high-frequency information. Because learning-based approaches have proven successful in the past in terms of addressing the super-resolution problem, we focus on a learning-based technique for compressed image denoising. In this process, it is important to search for the exact prior in a training set. The proposed method utilizes two different databases (i.e., a noisy and a denoised database), which work together in a complementary way. The denoised images from the dual databases are combined into a final denoised one. Additionally, the input noisy image is decomposed into structure and texture components, and only the latter is denoised because most noise tends to exist within the texture component. Experimental results show that the proposed method can reduce compression noise while reconstructing the original information that was lost in the compression process, especially for texture regions.

Original languageEnglish
Pages (from-to)98-107
Number of pages10
JournalJournal of Visual Communication and Image Representation
Volume43
DOIs
Publication statusPublished - 2017 Feb 1

Fingerprint

Noise abatement
Textures
Image denoising
Restoration

Keywords

  • Compression noise
  • Dual learning
  • Learning-based denoising
  • Texture domain

ASJC Scopus subject areas

  • Signal Processing
  • Media Technology
  • Computer Vision and Pattern Recognition
  • Electrical and Electronic Engineering

Cite this

Dual learning based compression noise reduction in the texture domain. / Lee, Jae Won; Lee, Oh Young; Kim, Jong-Ok.

In: Journal of Visual Communication and Image Representation, Vol. 43, 01.02.2017, p. 98-107.

Research output: Contribution to journalArticle

@article{23ea1eef0e1141729e19780aa0065cfd,
title = "Dual learning based compression noise reduction in the texture domain",
abstract = "Compression noise reduction is similar to the super-resolution problem in terms of the restoration of lost high-frequency information. Because learning-based approaches have proven successful in the past in terms of addressing the super-resolution problem, we focus on a learning-based technique for compressed image denoising. In this process, it is important to search for the exact prior in a training set. The proposed method utilizes two different databases (i.e., a noisy and a denoised database), which work together in a complementary way. The denoised images from the dual databases are combined into a final denoised one. Additionally, the input noisy image is decomposed into structure and texture components, and only the latter is denoised because most noise tends to exist within the texture component. Experimental results show that the proposed method can reduce compression noise while reconstructing the original information that was lost in the compression process, especially for texture regions.",
keywords = "Compression noise, Dual learning, Learning-based denoising, Texture domain",
author = "Lee, {Jae Won} and Lee, {Oh Young} and Jong-Ok Kim",
year = "2017",
month = "2",
day = "1",
doi = "10.1016/j.jvcir.2016.12.014",
language = "English",
volume = "43",
pages = "98--107",
journal = "Journal of Visual Communication and Image Representation",
issn = "1047-3203",
publisher = "Academic Press Inc.",

}

TY - JOUR

T1 - Dual learning based compression noise reduction in the texture domain

AU - Lee, Jae Won

AU - Lee, Oh Young

AU - Kim, Jong-Ok

PY - 2017/2/1

Y1 - 2017/2/1

N2 - Compression noise reduction is similar to the super-resolution problem in terms of the restoration of lost high-frequency information. Because learning-based approaches have proven successful in the past in terms of addressing the super-resolution problem, we focus on a learning-based technique for compressed image denoising. In this process, it is important to search for the exact prior in a training set. The proposed method utilizes two different databases (i.e., a noisy and a denoised database), which work together in a complementary way. The denoised images from the dual databases are combined into a final denoised one. Additionally, the input noisy image is decomposed into structure and texture components, and only the latter is denoised because most noise tends to exist within the texture component. Experimental results show that the proposed method can reduce compression noise while reconstructing the original information that was lost in the compression process, especially for texture regions.

AB - Compression noise reduction is similar to the super-resolution problem in terms of the restoration of lost high-frequency information. Because learning-based approaches have proven successful in the past in terms of addressing the super-resolution problem, we focus on a learning-based technique for compressed image denoising. In this process, it is important to search for the exact prior in a training set. The proposed method utilizes two different databases (i.e., a noisy and a denoised database), which work together in a complementary way. The denoised images from the dual databases are combined into a final denoised one. Additionally, the input noisy image is decomposed into structure and texture components, and only the latter is denoised because most noise tends to exist within the texture component. Experimental results show that the proposed method can reduce compression noise while reconstructing the original information that was lost in the compression process, especially for texture regions.

KW - Compression noise

KW - Dual learning

KW - Learning-based denoising

KW - Texture domain

UR - http://www.scopus.com/inward/record.url?scp=85008637619&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85008637619&partnerID=8YFLogxK

U2 - 10.1016/j.jvcir.2016.12.014

DO - 10.1016/j.jvcir.2016.12.014

M3 - Article

AN - SCOPUS:85008637619

VL - 43

SP - 98

EP - 107

JO - Journal of Visual Communication and Image Representation

JF - Journal of Visual Communication and Image Representation

SN - 1047-3203

ER -