Sparsity-Based Color Image Super Resolution via Exploiting Cross Channel Constraints

Hojjat Seyed Mousavi, Vishal Monga

Research output: Contribution to journalArticle

19 Citations (Scopus)

Abstract

Sparsity constrained single image super-resolution (SR) has been of much recent interest. A typical approach involves sparsely representing patches in a low-resolution (LR) input image via a dictionary of example LR patches, and then using the coefficients of this representation to generate the high-resolution (HR) output via an analogous HR dictionary. However, most existing sparse representation methods for SR focus on the luminance channel information and do not capture interactions between color channels. In this paper, we extend sparsity-based SR to multiple color channels by taking the color information into account. Edge similarities amongst RGB color bands are exploited as cross channel correlation constraints. These additional constraints lead to a new optimization problem, which is not easily solvable; however, a tractable solution is proposed to solve it efficiently. Moreover, to fully exploit the complementary information among color channels, a dictionary learning method is also proposed specifically to learn color dictionaries that encourage edge similarities. Merits of the proposed method over state of the art are demonstrated both visually and quantitatively using image quality metrics.

Original languageEnglish (US)
Article number7929299
Pages (from-to)5094-5106
Number of pages13
JournalIEEE Transactions on Image Processing
Volume26
Issue number11
DOIs
StatePublished - Nov 2017

Fingerprint

Glossaries
Color
Image quality
Luminance

All Science Journal Classification (ASJC) codes

  • Software
  • Computer Graphics and Computer-Aided Design

Cite this

@article{b282fec4d7e6442982f903a548f16f78,
title = "Sparsity-Based Color Image Super Resolution via Exploiting Cross Channel Constraints",
abstract = "Sparsity constrained single image super-resolution (SR) has been of much recent interest. A typical approach involves sparsely representing patches in a low-resolution (LR) input image via a dictionary of example LR patches, and then using the coefficients of this representation to generate the high-resolution (HR) output via an analogous HR dictionary. However, most existing sparse representation methods for SR focus on the luminance channel information and do not capture interactions between color channels. In this paper, we extend sparsity-based SR to multiple color channels by taking the color information into account. Edge similarities amongst RGB color bands are exploited as cross channel correlation constraints. These additional constraints lead to a new optimization problem, which is not easily solvable; however, a tractable solution is proposed to solve it efficiently. Moreover, to fully exploit the complementary information among color channels, a dictionary learning method is also proposed specifically to learn color dictionaries that encourage edge similarities. Merits of the proposed method over state of the art are demonstrated both visually and quantitatively using image quality metrics.",
author = "Mousavi, {Hojjat Seyed} and Vishal Monga",
year = "2017",
month = "11",
doi = "10.1109/TIP.2017.2704443",
language = "English (US)",
volume = "26",
pages = "5094--5106",
journal = "IEEE Transactions on Image Processing",
issn = "1057-7149",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "11",

}

Sparsity-Based Color Image Super Resolution via Exploiting Cross Channel Constraints. / Mousavi, Hojjat Seyed; Monga, Vishal.

In: IEEE Transactions on Image Processing, Vol. 26, No. 11, 7929299, 11.2017, p. 5094-5106.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Sparsity-Based Color Image Super Resolution via Exploiting Cross Channel Constraints

AU - Mousavi, Hojjat Seyed

AU - Monga, Vishal

PY - 2017/11

Y1 - 2017/11

N2 - Sparsity constrained single image super-resolution (SR) has been of much recent interest. A typical approach involves sparsely representing patches in a low-resolution (LR) input image via a dictionary of example LR patches, and then using the coefficients of this representation to generate the high-resolution (HR) output via an analogous HR dictionary. However, most existing sparse representation methods for SR focus on the luminance channel information and do not capture interactions between color channels. In this paper, we extend sparsity-based SR to multiple color channels by taking the color information into account. Edge similarities amongst RGB color bands are exploited as cross channel correlation constraints. These additional constraints lead to a new optimization problem, which is not easily solvable; however, a tractable solution is proposed to solve it efficiently. Moreover, to fully exploit the complementary information among color channels, a dictionary learning method is also proposed specifically to learn color dictionaries that encourage edge similarities. Merits of the proposed method over state of the art are demonstrated both visually and quantitatively using image quality metrics.

AB - Sparsity constrained single image super-resolution (SR) has been of much recent interest. A typical approach involves sparsely representing patches in a low-resolution (LR) input image via a dictionary of example LR patches, and then using the coefficients of this representation to generate the high-resolution (HR) output via an analogous HR dictionary. However, most existing sparse representation methods for SR focus on the luminance channel information and do not capture interactions between color channels. In this paper, we extend sparsity-based SR to multiple color channels by taking the color information into account. Edge similarities amongst RGB color bands are exploited as cross channel correlation constraints. These additional constraints lead to a new optimization problem, which is not easily solvable; however, a tractable solution is proposed to solve it efficiently. Moreover, to fully exploit the complementary information among color channels, a dictionary learning method is also proposed specifically to learn color dictionaries that encourage edge similarities. Merits of the proposed method over state of the art are demonstrated both visually and quantitatively using image quality metrics.

UR - http://www.scopus.com/inward/record.url?scp=85029545930&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85029545930&partnerID=8YFLogxK

U2 - 10.1109/TIP.2017.2704443

DO - 10.1109/TIP.2017.2704443

M3 - Article

C2 - 28534773

AN - SCOPUS:85029545930

VL - 26

SP - 5094

EP - 5106

JO - IEEE Transactions on Image Processing

JF - IEEE Transactions on Image Processing

SN - 1057-7149

IS - 11

M1 - 7929299

ER -