An image quality assessment algorithm based on feature selection

Ting Lu, Yanning Zhang, Haisen Li

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

Image Quality Assessment(IQA) is of fundamental importance to numerous imaging and video processing applications. For most of the applications, the perceptual meaningful measure is the one which can automatically assess the quality of images or videos in a perceptually consistent manner. However, most commonly used IQA metrics are not consistent well with the human judgments of image quality. Recently, the SSIM metric which takes people's visual characteristics into consideration performs much better than the traditional PSNR/MSE. But the defects of it still exit on some specific kinds of distortions. A new algorithm of IQA based on feature selection is proposed in this paper. Local gradient entropy and phase congruency are added to the SSIM framework. Through in-depth feature selection and definition plus better pooling strategy, this algorithm performs much better in LIVE datasets.

Original languageEnglish
Title of host publicationIntelligent Science and Intelligent Data Engineering - Third Sino-Foreign-Interchange Workshop, IScIDE 2012, Revised Selected Papers
Pages289-297
Number of pages9
DOIs
StatePublished - 2013
Event3rd Sino-Foreign-Interchange Workshop on Intelligent Science and Intelligent Data Engineering, IScIDE 2012 - Nanjing, China
Duration: 15 Oct 201217 Oct 2012

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume7751 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference3rd Sino-Foreign-Interchange Workshop on Intelligent Science and Intelligent Data Engineering, IScIDE 2012
Country/TerritoryChina
CityNanjing
Period15/10/1217/10/12

Keywords

  • Feature selection
  • Image quality assessment
  • Local gradient entropy
  • Phase congruency

Fingerprint

Dive into the research topics of 'An image quality assessment algorithm based on feature selection'. Together they form a unique fingerprint.

Cite this