A New Full-Reference Image Quality Metric for Motion Blur Profile Characterization

Motion blur is common in images captured by handheld devices, arising from hand, device and/or object motion. To restore sharp images from the images degraded by the motion, it is extremely important to assess the quality of the captured image and its corresponding blur profile as accurately as poss...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Mohammad Abdullah-Al-Mamun, Vivek Tyagi, Hong Zhao
Formato:	article
Lenguaje:	EN
Publicado:	IEEE 2021
Materias:	Blur extent blur kernel blur level blur metric full reference image quality assessment motion blur Electrical engineering. Electronics. Nuclear engineering TK1-9971
Acceso en línea:	https://doaj.org/article/fff899ced8144bc1a76639d3e4dbe2c9
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:fff899ced8144bc1a76639d3e4dbe2c9
record_format	dspace
spelling	oai:doaj.org-article:fff899ced8144bc1a76639d3e4dbe2c92021-12-02T00:00:19ZA New Full-Reference Image Quality Metric for Motion Blur Profile Characterization2169-353610.1109/ACCESS.2021.3130177https://doaj.org/article/fff899ced8144bc1a76639d3e4dbe2c92021-01-01T00:00:00Zhttps://ieeexplore.ieee.org/document/9624980/https://doaj.org/toc/2169-3536Motion blur is common in images captured by handheld devices, arising from hand, device and/or object motion. To restore sharp images from the images degraded by the motion, it is extremely important to assess the quality of the captured image and its corresponding blur profile as accurately as possible. In image deblurring, the perceived image quality is usually assessed by the SSIM and the PSNR metric. These methods have certain limitations and the objective image quality assessed by these methods can be contradictory to the subjectively perceived image quality. We propose a new reference image based objective blur level (BL) metric by utilizing point spread function/blur kernel analysis in this paper. In our experiments, we found our BL metric describes the perceived image quality of motion blurred images better than SSIM and PSNR in most cases. Additionally, our method performs well in low light and low texture images, where SSIM and PSNR metrics are prone to failure in describing blurriness/sharpness of the image.Mohammad Abdullah-Al-MamunVivek TyagiHong ZhaoIEEEarticleBlur extentblur kernelblur levelblur metricfull reference image quality assessmentmotion blurElectrical engineering. Electronics. Nuclear engineeringTK1-9971ENIEEE Access, Vol 9, Pp 156361-156371 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Blur extent blur kernel blur level blur metric full reference image quality assessment motion blur Electrical engineering. Electronics. Nuclear engineering TK1-9971
spellingShingle	Blur extent blur kernel blur level blur metric full reference image quality assessment motion blur Electrical engineering. Electronics. Nuclear engineering TK1-9971 Mohammad Abdullah-Al-Mamun Vivek Tyagi Hong Zhao A New Full-Reference Image Quality Metric for Motion Blur Profile Characterization
description	Motion blur is common in images captured by handheld devices, arising from hand, device and/or object motion. To restore sharp images from the images degraded by the motion, it is extremely important to assess the quality of the captured image and its corresponding blur profile as accurately as possible. In image deblurring, the perceived image quality is usually assessed by the SSIM and the PSNR metric. These methods have certain limitations and the objective image quality assessed by these methods can be contradictory to the subjectively perceived image quality. We propose a new reference image based objective blur level (BL) metric by utilizing point spread function/blur kernel analysis in this paper. In our experiments, we found our BL metric describes the perceived image quality of motion blurred images better than SSIM and PSNR in most cases. Additionally, our method performs well in low light and low texture images, where SSIM and PSNR metrics are prone to failure in describing blurriness/sharpness of the image.
format	article
author	Mohammad Abdullah-Al-Mamun Vivek Tyagi Hong Zhao
author_facet	Mohammad Abdullah-Al-Mamun Vivek Tyagi Hong Zhao
author_sort	Mohammad Abdullah-Al-Mamun
title	A New Full-Reference Image Quality Metric for Motion Blur Profile Characterization
title_short	A New Full-Reference Image Quality Metric for Motion Blur Profile Characterization
title_full	A New Full-Reference Image Quality Metric for Motion Blur Profile Characterization
title_fullStr	A New Full-Reference Image Quality Metric for Motion Blur Profile Characterization
title_full_unstemmed	A New Full-Reference Image Quality Metric for Motion Blur Profile Characterization
title_sort	new full-reference image quality metric for motion blur profile characterization
publisher	IEEE
publishDate	2021
url	https://doaj.org/article/fff899ced8144bc1a76639d3e4dbe2c9
work_keys_str_mv	AT mohammadabdullahalmamun anewfullreferenceimagequalitymetricformotionblurprofilecharacterization AT vivektyagi anewfullreferenceimagequalitymetricformotionblurprofilecharacterization AT hongzhao anewfullreferenceimagequalitymetricformotionblurprofilecharacterization AT mohammadabdullahalmamun newfullreferenceimagequalitymetricformotionblurprofilecharacterization AT vivektyagi newfullreferenceimagequalitymetricformotionblurprofilecharacterization AT hongzhao newfullreferenceimagequalitymetricformotionblurprofilecharacterization
_version_	1718404022148268032

A New Full-Reference Image Quality Metric for Motion Blur Profile Characterization

Ejemplares similares