No-reference image quality assessment for contrastdistorted images using statistical features in Curvelet domain

Ismail T. Ahmed, Soong Der Chen

Research output: Contribution to journalArticle

Abstract

Most No-Reference Image Quality Assessment (NR-IQA) metrics are designed for the quality assessment of images distorted by compression, noise and blurring. Few NR-IQA metrics exist for Contrast-Distorted Images (CDI).Reduced-reference Image Quality Metric for Contrast-changed images (RIQMC) and NR-IQA for Contrast- Distorted Images (NR-IQACDI) are the state-of-the-art IQA algorithms for CDI. Room for improvement exists, especially for the assessment results using the image database called TID2013. The current NR-IQACDI uses features in spatial domain. This paper proposes the use of the same statistical features but in Curvelet domain, which is powerful in capturing the multiscale and multidirectional information of an image. Experiments are conducted to assess the effect of using statistical features in Curvelet domain. The experiment results are based on K-fold cross validation with K range from (2 to 10).The statistical tests indicate that the performance using selected statistical features in the Curvelet domain are better than that of the NRIQACDI. The use of other statistical features and selection methods should be further investigated to increase the prediction performance.

Original languageEnglish
Pages (from-to)3613-3620
Number of pages8
JournalARPN Journal of Engineering and Applied Sciences
Volume12
Issue number11
Publication statusPublished - 01 Jun 2017

Fingerprint

Image quality
Statistical tests
Experiments

All Science Journal Classification (ASJC) codes

  • Engineering(all)

Cite this

@article{52a64e024f3c432aae098d3794cb658a,
title = "No-reference image quality assessment for contrastdistorted images using statistical features in Curvelet domain",
abstract = "Most No-Reference Image Quality Assessment (NR-IQA) metrics are designed for the quality assessment of images distorted by compression, noise and blurring. Few NR-IQA metrics exist for Contrast-Distorted Images (CDI).Reduced-reference Image Quality Metric for Contrast-changed images (RIQMC) and NR-IQA for Contrast- Distorted Images (NR-IQACDI) are the state-of-the-art IQA algorithms for CDI. Room for improvement exists, especially for the assessment results using the image database called TID2013. The current NR-IQACDI uses features in spatial domain. This paper proposes the use of the same statistical features but in Curvelet domain, which is powerful in capturing the multiscale and multidirectional information of an image. Experiments are conducted to assess the effect of using statistical features in Curvelet domain. The experiment results are based on K-fold cross validation with K range from (2 to 10).The statistical tests indicate that the performance using selected statistical features in the Curvelet domain are better than that of the NRIQACDI. The use of other statistical features and selection methods should be further investigated to increase the prediction performance.",
author = "Ahmed, {Ismail T.} and Chen, {Soong Der}",
year = "2017",
month = "6",
day = "1",
language = "English",
volume = "12",
pages = "3613--3620",
journal = "ARPN Journal of Engineering and Applied Sciences",
issn = "1819-6608",
publisher = "Asian Research Publishing Network (ARPN)",
number = "11",

}

No-reference image quality assessment for contrastdistorted images using statistical features in Curvelet domain. / Ahmed, Ismail T.; Chen, Soong Der.

In: ARPN Journal of Engineering and Applied Sciences, Vol. 12, No. 11, 01.06.2017, p. 3613-3620.

Research output: Contribution to journalArticle

TY - JOUR

T1 - No-reference image quality assessment for contrastdistorted images using statistical features in Curvelet domain

AU - Ahmed, Ismail T.

AU - Chen, Soong Der

PY - 2017/6/1

Y1 - 2017/6/1

N2 - Most No-Reference Image Quality Assessment (NR-IQA) metrics are designed for the quality assessment of images distorted by compression, noise and blurring. Few NR-IQA metrics exist for Contrast-Distorted Images (CDI).Reduced-reference Image Quality Metric for Contrast-changed images (RIQMC) and NR-IQA for Contrast- Distorted Images (NR-IQACDI) are the state-of-the-art IQA algorithms for CDI. Room for improvement exists, especially for the assessment results using the image database called TID2013. The current NR-IQACDI uses features in spatial domain. This paper proposes the use of the same statistical features but in Curvelet domain, which is powerful in capturing the multiscale and multidirectional information of an image. Experiments are conducted to assess the effect of using statistical features in Curvelet domain. The experiment results are based on K-fold cross validation with K range from (2 to 10).The statistical tests indicate that the performance using selected statistical features in the Curvelet domain are better than that of the NRIQACDI. The use of other statistical features and selection methods should be further investigated to increase the prediction performance.

AB - Most No-Reference Image Quality Assessment (NR-IQA) metrics are designed for the quality assessment of images distorted by compression, noise and blurring. Few NR-IQA metrics exist for Contrast-Distorted Images (CDI).Reduced-reference Image Quality Metric for Contrast-changed images (RIQMC) and NR-IQA for Contrast- Distorted Images (NR-IQACDI) are the state-of-the-art IQA algorithms for CDI. Room for improvement exists, especially for the assessment results using the image database called TID2013. The current NR-IQACDI uses features in spatial domain. This paper proposes the use of the same statistical features but in Curvelet domain, which is powerful in capturing the multiscale and multidirectional information of an image. Experiments are conducted to assess the effect of using statistical features in Curvelet domain. The experiment results are based on K-fold cross validation with K range from (2 to 10).The statistical tests indicate that the performance using selected statistical features in the Curvelet domain are better than that of the NRIQACDI. The use of other statistical features and selection methods should be further investigated to increase the prediction performance.

UR - http://www.scopus.com/inward/record.url?scp=85020542408&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85020542408&partnerID=8YFLogxK

M3 - Article

VL - 12

SP - 3613

EP - 3620

JO - ARPN Journal of Engineering and Applied Sciences

JF - ARPN Journal of Engineering and Applied Sciences

SN - 1819-6608

IS - 11

ER -