Illumination normalization for edge-based face recognition using the fusion of RGB normalization and gamma correction

Chollette C. Chude-Olisah, Ghazali Sulong, Uche A.K. Chude-Okonkwo, Siti Z.M. Hashim

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

19 Citations (Scopus)

Abstract

In this paper, an illumination normalization technique for edge-based face recognition on face images with non-uniform illumination conditions, is proposed. The proposed illumination normalization technique fuses the merits of color (Red, Green and Blue) normalization (Nrgb) and gamma correction (GC) for color images. By the fusion of these methods the image becomes independent of the change in face images due to illumination direction. In that way, the presence of false edges in gradient faces is reduced. Experimental results on Georgia Tech Face database with illumination problem shows that the proposed technique improved significantly recognition accuracy in comparison to histogram equalization (HE), logarithm transform (LT) and gamma correction (GC).

Original languageEnglish
Title of host publicationIEEE ICSIPA 2013 - IEEE International Conference on Signal and Image Processing Applications
PublisherIEEE Computer Society
Pages412-416
Number of pages5
ISBN (Print)9781479902675
DOIs
Publication statusPublished - 2013
Externally publishedYes
Event2013 3rd IEEE International Conference on Signal and Image Processing Applications, IEEE ICSIPA 2013 - Melaka, Malaysia
Duration: 8 Oct 201310 Oct 2013

Publication series

NameIEEE ICSIPA 2013 - IEEE International Conference on Signal and Image Processing Applications

Conference

Conference2013 3rd IEEE International Conference on Signal and Image Processing Applications, IEEE ICSIPA 2013
Country/TerritoryMalaysia
CityMelaka
Period8/10/1310/10/13

Keywords

  • RGB normalization
  • edge-based face recognition
  • gamma color correction
  • illumination normalization
  • non-uniform illumination

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Computer Vision and Pattern Recognition

Fingerprint

Dive into the research topics of 'Illumination normalization for edge-based face recognition using the fusion of RGB normalization and gamma correction'. Together they form a unique fingerprint.

Cite this