Skip to main navigation Skip to search Skip to main content

A novel metric for image denoising algorithms

  • School of Computer Science and Technology, Harbin Institute of Technology
  • Utah State University

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

Abstract

Denoising algorithms, especially, the ones with contrast enhancement capability have many important applications. However, there is no an effective and accurate measurement for evaluating their performance objectively. This introduces a new metric, HME (Homogeneity Mean Error), to assess the denoising algorithms, especially those with enhancement capability. HME is based on the homogeneity property of each pixel which is sensitive to the changes of the structural information and noise levels, but insensitive to the changes of the contrast. Therefore, it can be utilized for evaluating the denoising algorithms. Various experiments are performed on images corrupted with different type of noise, the results demonstrate that HME is an effective and accurate metric for assessing the denoising algorithms with/without contrast enhancement. .

Original languageEnglish
Title of host publicationIntelligence Science and Big Data Engineering - 4th International Conference, IScIDE 2013, Revised Selected Papers
PublisherSpringer Verlag
Pages538-545
Number of pages8
ISBN (Print)9783642420566
DOIs
StatePublished - 2013
Externally publishedYes
Event4th International Conference on Intelligence Science and Big Data Engineering, IScIDE 2013 - Beijing, China
Duration: 31 Jul 20132 Aug 2013

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume8261 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference4th International Conference on Intelligence Science and Big Data Engineering, IScIDE 2013
Country/TerritoryChina
CityBeijing
Period31/07/132/08/13

Keywords

  • HMD (Homogeneity Mean Difference)
  • contrast enhancement
  • denoising performance
  • image quality assessment (QA)
  • objective criterion

Fingerprint

Dive into the research topics of 'A novel metric for image denoising algorithms'. Together they form a unique fingerprint.

Cite this