An improved method for illumination invariant face recognition based on adaptive rescaling DCT coefficient in logarithm domain

Chao Yu, Xiaoqun Zhao, Seiichiro Kamata

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper presents an improved method for robust face recognition using illumination normalization based on Discrete Cosine Transform (DCT) in logarithm domain. Two novel coefficients are designed to identify the lighting condition (LC), based on which the low-frequency DCT coefficients are adaptively rescaled except the first one (DC). As a result variations under different illumination conditions are minimized meanwhile original information contained in low-frequency is comparatively well preserved. Results of experiments on Yale B database and Extended Yale B database show that proposed method has better performance under variational input illumination conditions. The proposed method is fast in computation and could be easily implemented into real time face recognition systems.

Original languageEnglish
Title of host publicationLecture Notes in Electrical Engineering
Pages297-304
Number of pages8
Volume121 LNEE
DOIs
Publication statusPublished - 2011
Event2011 International Conference on Computer, Communication, Control and Automation, 3CA 2011 - Zhuhai
Duration: 2011 Nov 192011 Nov 20

Publication series

NameLecture Notes in Electrical Engineering
Volume121 LNEE
ISSN (Print)18761100
ISSN (Electronic)18761119

Other

Other2011 International Conference on Computer, Communication, Control and Automation, 3CA 2011
CityZhuhai
Period11/11/1911/11/20

Fingerprint

Discrete cosine transforms
Face recognition
Lighting
Experiments

Keywords

  • DCT
  • face recognition
  • illumination invariant
  • LC coefficient

ASJC Scopus subject areas

  • Industrial and Manufacturing Engineering

Cite this

Yu, C., Zhao, X., & Kamata, S. (2011). An improved method for illumination invariant face recognition based on adaptive rescaling DCT coefficient in logarithm domain. In Lecture Notes in Electrical Engineering (Vol. 121 LNEE, pp. 297-304). (Lecture Notes in Electrical Engineering; Vol. 121 LNEE). https://doi.org/10.1007/978-3-642-25541-0_39

An improved method for illumination invariant face recognition based on adaptive rescaling DCT coefficient in logarithm domain. / Yu, Chao; Zhao, Xiaoqun; Kamata, Seiichiro.

Lecture Notes in Electrical Engineering. Vol. 121 LNEE 2011. p. 297-304 (Lecture Notes in Electrical Engineering; Vol. 121 LNEE).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Yu, C, Zhao, X & Kamata, S 2011, An improved method for illumination invariant face recognition based on adaptive rescaling DCT coefficient in logarithm domain. in Lecture Notes in Electrical Engineering. vol. 121 LNEE, Lecture Notes in Electrical Engineering, vol. 121 LNEE, pp. 297-304, 2011 International Conference on Computer, Communication, Control and Automation, 3CA 2011, Zhuhai, 11/11/19. https://doi.org/10.1007/978-3-642-25541-0_39
Yu C, Zhao X, Kamata S. An improved method for illumination invariant face recognition based on adaptive rescaling DCT coefficient in logarithm domain. In Lecture Notes in Electrical Engineering. Vol. 121 LNEE. 2011. p. 297-304. (Lecture Notes in Electrical Engineering). https://doi.org/10.1007/978-3-642-25541-0_39
Yu, Chao ; Zhao, Xiaoqun ; Kamata, Seiichiro. / An improved method for illumination invariant face recognition based on adaptive rescaling DCT coefficient in logarithm domain. Lecture Notes in Electrical Engineering. Vol. 121 LNEE 2011. pp. 297-304 (Lecture Notes in Electrical Engineering).
@inproceedings{9fc196a511fd4a8f93b9fbde7b259e61,
title = "An improved method for illumination invariant face recognition based on adaptive rescaling DCT coefficient in logarithm domain",
abstract = "This paper presents an improved method for robust face recognition using illumination normalization based on Discrete Cosine Transform (DCT) in logarithm domain. Two novel coefficients are designed to identify the lighting condition (LC), based on which the low-frequency DCT coefficients are adaptively rescaled except the first one (DC). As a result variations under different illumination conditions are minimized meanwhile original information contained in low-frequency is comparatively well preserved. Results of experiments on Yale B database and Extended Yale B database show that proposed method has better performance under variational input illumination conditions. The proposed method is fast in computation and could be easily implemented into real time face recognition systems.",
keywords = "DCT, face recognition, illumination invariant, LC coefficient",
author = "Chao Yu and Xiaoqun Zhao and Seiichiro Kamata",
year = "2011",
doi = "10.1007/978-3-642-25541-0_39",
language = "English",
isbn = "9783642255403",
volume = "121 LNEE",
series = "Lecture Notes in Electrical Engineering",
pages = "297--304",
booktitle = "Lecture Notes in Electrical Engineering",

}

TY - GEN

T1 - An improved method for illumination invariant face recognition based on adaptive rescaling DCT coefficient in logarithm domain

AU - Yu, Chao

AU - Zhao, Xiaoqun

AU - Kamata, Seiichiro

PY - 2011

Y1 - 2011

N2 - This paper presents an improved method for robust face recognition using illumination normalization based on Discrete Cosine Transform (DCT) in logarithm domain. Two novel coefficients are designed to identify the lighting condition (LC), based on which the low-frequency DCT coefficients are adaptively rescaled except the first one (DC). As a result variations under different illumination conditions are minimized meanwhile original information contained in low-frequency is comparatively well preserved. Results of experiments on Yale B database and Extended Yale B database show that proposed method has better performance under variational input illumination conditions. The proposed method is fast in computation and could be easily implemented into real time face recognition systems.

AB - This paper presents an improved method for robust face recognition using illumination normalization based on Discrete Cosine Transform (DCT) in logarithm domain. Two novel coefficients are designed to identify the lighting condition (LC), based on which the low-frequency DCT coefficients are adaptively rescaled except the first one (DC). As a result variations under different illumination conditions are minimized meanwhile original information contained in low-frequency is comparatively well preserved. Results of experiments on Yale B database and Extended Yale B database show that proposed method has better performance under variational input illumination conditions. The proposed method is fast in computation and could be easily implemented into real time face recognition systems.

KW - DCT

KW - face recognition

KW - illumination invariant

KW - LC coefficient

UR - http://www.scopus.com/inward/record.url?scp=82255165274&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=82255165274&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-25541-0_39

DO - 10.1007/978-3-642-25541-0_39

M3 - Conference contribution

AN - SCOPUS:82255165274

SN - 9783642255403

VL - 121 LNEE

T3 - Lecture Notes in Electrical Engineering

SP - 297

EP - 304

BT - Lecture Notes in Electrical Engineering

ER -