Statistical evaluation of image quality measures

Ismail Avcibaş, Bülent Sankur, Khalid Sayood

Research output: Contribution to journalArticle

425 Citations (Scopus)

Abstract

In this work we comprehensively categorize image quality measures, extend measures defined for gray scale images to their multispectral case, and propose novel image quality measures. They are categorized into pixel difference-based, correlation-based, edge-based, spectral-based, context-based and human visual system (HVS)-based measures. Furthermore we compare these measures statistically for still image compression applications. The statistical behavior of the measures and their sensitivity to coding artifacts are investigated via analysis of variance techniques. Their similarities or differences are illustrated by plotting their Kohonen maps. Measures that give consistent scores across an image class and that are sensitive to coding artifacts are pointed out. It was found that measures based on the phase spectrum, the multiresolution distance or the HVS filtered mean square error are computationally simple and are more responsive to coding artifacts. We also demonstrate the utility of combining selected quality metrics in building a steganalysis tool.

Original languageEnglish (US)
Pages (from-to)206-223
Number of pages18
JournalJournal of Electronic Imaging
Volume11
Issue number2
DOIs
StatePublished - Jan 1 2002

Fingerprint

Image quality
evaluation
Self organizing maps
Analysis of variance (ANOVA)
Image compression
Mean square error
Pixels
artifacts
coding
analysis of variance
gray scale
plotting
pixels
sensitivity

ASJC Scopus subject areas

  • Atomic and Molecular Physics, and Optics
  • Computer Science Applications
  • Electrical and Electronic Engineering

Cite this

Statistical evaluation of image quality measures. / Avcibaş, Ismail; Sankur, Bülent; Sayood, Khalid.

In: Journal of Electronic Imaging, Vol. 11, No. 2, 01.01.2002, p. 206-223.

Research output: Contribution to journalArticle

Avcibaş, Ismail ; Sankur, Bülent ; Sayood, Khalid. / Statistical evaluation of image quality measures. In: Journal of Electronic Imaging. 2002 ; Vol. 11, No. 2. pp. 206-223.
@article{55bf346e4ce349648f03c56892a1339c,
title = "Statistical evaluation of image quality measures",
abstract = "In this work we comprehensively categorize image quality measures, extend measures defined for gray scale images to their multispectral case, and propose novel image quality measures. They are categorized into pixel difference-based, correlation-based, edge-based, spectral-based, context-based and human visual system (HVS)-based measures. Furthermore we compare these measures statistically for still image compression applications. The statistical behavior of the measures and their sensitivity to coding artifacts are investigated via analysis of variance techniques. Their similarities or differences are illustrated by plotting their Kohonen maps. Measures that give consistent scores across an image class and that are sensitive to coding artifacts are pointed out. It was found that measures based on the phase spectrum, the multiresolution distance or the HVS filtered mean square error are computationally simple and are more responsive to coding artifacts. We also demonstrate the utility of combining selected quality metrics in building a steganalysis tool.",
author = "Ismail Avcibaş and B{\"u}lent Sankur and Khalid Sayood",
year = "2002",
month = "1",
day = "1",
doi = "10.1117/1.1455011",
language = "English (US)",
volume = "11",
pages = "206--223",
journal = "Journal of Electronic Imaging",
issn = "1017-9909",
publisher = "SPIE",
number = "2",

}

TY - JOUR

T1 - Statistical evaluation of image quality measures

AU - Avcibaş, Ismail

AU - Sankur, Bülent

AU - Sayood, Khalid

PY - 2002/1/1

Y1 - 2002/1/1

N2 - In this work we comprehensively categorize image quality measures, extend measures defined for gray scale images to their multispectral case, and propose novel image quality measures. They are categorized into pixel difference-based, correlation-based, edge-based, spectral-based, context-based and human visual system (HVS)-based measures. Furthermore we compare these measures statistically for still image compression applications. The statistical behavior of the measures and their sensitivity to coding artifacts are investigated via analysis of variance techniques. Their similarities or differences are illustrated by plotting their Kohonen maps. Measures that give consistent scores across an image class and that are sensitive to coding artifacts are pointed out. It was found that measures based on the phase spectrum, the multiresolution distance or the HVS filtered mean square error are computationally simple and are more responsive to coding artifacts. We also demonstrate the utility of combining selected quality metrics in building a steganalysis tool.

AB - In this work we comprehensively categorize image quality measures, extend measures defined for gray scale images to their multispectral case, and propose novel image quality measures. They are categorized into pixel difference-based, correlation-based, edge-based, spectral-based, context-based and human visual system (HVS)-based measures. Furthermore we compare these measures statistically for still image compression applications. The statistical behavior of the measures and their sensitivity to coding artifacts are investigated via analysis of variance techniques. Their similarities or differences are illustrated by plotting their Kohonen maps. Measures that give consistent scores across an image class and that are sensitive to coding artifacts are pointed out. It was found that measures based on the phase spectrum, the multiresolution distance or the HVS filtered mean square error are computationally simple and are more responsive to coding artifacts. We also demonstrate the utility of combining selected quality metrics in building a steganalysis tool.

UR - http://www.scopus.com/inward/record.url?scp=0036530067&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0036530067&partnerID=8YFLogxK

U2 - 10.1117/1.1455011

DO - 10.1117/1.1455011

M3 - Article

VL - 11

SP - 206

EP - 223

JO - Journal of Electronic Imaging

JF - Journal of Electronic Imaging

SN - 1017-9909

IS - 2

ER -