Cargando…

Effective Three-Stage Demosaicking Method for RGBW CFA Images Using The Iterative Error-Compensation Based Approach

As the color filter array (CFA)2.0, the RGBW CFA pattern, in which each CFA pixel contains only one R, G, B, or W color value, provides more luminance information than the Bayer CFA pattern. Demosaicking RGBW CFA images [Formula: see text] is necessary in order to provide high-quality RGB full-color...

Descripción completa

Detalles Bibliográficos
Autores principales: Chung, Kuo-Liang, Chan, Tzu-Hsien, Chen, Szu-Ni
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2020
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7412501/
https://www.ncbi.nlm.nih.gov/pubmed/32674284
http://dx.doi.org/10.3390/s20143908
_version_ 1783568622906507264
author Chung, Kuo-Liang
Chan, Tzu-Hsien
Chen, Szu-Ni
author_facet Chung, Kuo-Liang
Chan, Tzu-Hsien
Chen, Szu-Ni
author_sort Chung, Kuo-Liang
collection PubMed
description As the color filter array (CFA)2.0, the RGBW CFA pattern, in which each CFA pixel contains only one R, G, B, or W color value, provides more luminance information than the Bayer CFA pattern. Demosaicking RGBW CFA images [Formula: see text] is necessary in order to provide high-quality RGB full-color images as the target images for human perception. In this letter, we propose a three-stage demosaicking method for [Formula: see text]. In the first-stage, a cross shape-based color difference approach is proposed in order to interpolate the missing W color pixels in the W color plane of [Formula: see text]. In the second stage, an iterative error compensation-based demosaicking process is proposed to improve the quality of the demosaiced RGB full-color image. In the third stage, taking the input image [Formula: see text] as the ground truth RGBW CFA image, an [Formula: see text]-based refinement process is proposed to refine the quality of the demosaiced image obtained by the second stage. Based on the testing RGBW images that were collected from the Kodak and IMAX datasets, the comprehensive experimental results illustrated that the proposed three-stage demosaicking method achieves substantial quality and perceptual effect improvement relative to the previous method by Hamilton and Compton and the two state-of-the-art methods, Kwan et al.’s pansharpening-based method, and Kwan and Chou’s deep learning-based method.
format Online
Article
Text
id pubmed-7412501
institution National Center for Biotechnology Information
language English
publishDate 2020
publisher MDPI
record_format MEDLINE/PubMed
spelling pubmed-74125012020-08-26 Effective Three-Stage Demosaicking Method for RGBW CFA Images Using The Iterative Error-Compensation Based Approach Chung, Kuo-Liang Chan, Tzu-Hsien Chen, Szu-Ni Sensors (Basel) Letter As the color filter array (CFA)2.0, the RGBW CFA pattern, in which each CFA pixel contains only one R, G, B, or W color value, provides more luminance information than the Bayer CFA pattern. Demosaicking RGBW CFA images [Formula: see text] is necessary in order to provide high-quality RGB full-color images as the target images for human perception. In this letter, we propose a three-stage demosaicking method for [Formula: see text]. In the first-stage, a cross shape-based color difference approach is proposed in order to interpolate the missing W color pixels in the W color plane of [Formula: see text]. In the second stage, an iterative error compensation-based demosaicking process is proposed to improve the quality of the demosaiced RGB full-color image. In the third stage, taking the input image [Formula: see text] as the ground truth RGBW CFA image, an [Formula: see text]-based refinement process is proposed to refine the quality of the demosaiced image obtained by the second stage. Based on the testing RGBW images that were collected from the Kodak and IMAX datasets, the comprehensive experimental results illustrated that the proposed three-stage demosaicking method achieves substantial quality and perceptual effect improvement relative to the previous method by Hamilton and Compton and the two state-of-the-art methods, Kwan et al.’s pansharpening-based method, and Kwan and Chou’s deep learning-based method. MDPI 2020-07-14 /pmc/articles/PMC7412501/ /pubmed/32674284 http://dx.doi.org/10.3390/s20143908 Text en © 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
spellingShingle Letter
Chung, Kuo-Liang
Chan, Tzu-Hsien
Chen, Szu-Ni
Effective Three-Stage Demosaicking Method for RGBW CFA Images Using The Iterative Error-Compensation Based Approach
title Effective Three-Stage Demosaicking Method for RGBW CFA Images Using The Iterative Error-Compensation Based Approach
title_full Effective Three-Stage Demosaicking Method for RGBW CFA Images Using The Iterative Error-Compensation Based Approach
title_fullStr Effective Three-Stage Demosaicking Method for RGBW CFA Images Using The Iterative Error-Compensation Based Approach
title_full_unstemmed Effective Three-Stage Demosaicking Method for RGBW CFA Images Using The Iterative Error-Compensation Based Approach
title_short Effective Three-Stage Demosaicking Method for RGBW CFA Images Using The Iterative Error-Compensation Based Approach
title_sort effective three-stage demosaicking method for rgbw cfa images using the iterative error-compensation based approach
topic Letter
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7412501/
https://www.ncbi.nlm.nih.gov/pubmed/32674284
http://dx.doi.org/10.3390/s20143908
work_keys_str_mv AT chungkuoliang effectivethreestagedemosaickingmethodforrgbwcfaimagesusingtheiterativeerrorcompensationbasedapproach
AT chantzuhsien effectivethreestagedemosaickingmethodforrgbwcfaimagesusingtheiterativeerrorcompensationbasedapproach
AT chenszuni effectivethreestagedemosaickingmethodforrgbwcfaimagesusingtheiterativeerrorcompensationbasedapproach