Bidimensional empirical mode decomposition-based unlighting for face recognition

Miguel A. Ochoa-Villegas; Juan A. Nolazco-Flores; Olivia Barron-Cano; Ioannis A. Kakadiaris

doi:10.1109/WIFS.2014.7084297

Bidimensional empirical mode decomposition-based unlighting for face recognition

Miguel A. Ochoa-Villegas, Juan A. Nolazco-Flores, Olivia Barron-Cano, Ioannis A. Kakadiaris

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

1 Scopus citations

Abstract

A face recognition system must be capable of handling facial data with head pose variations or different illumination conditions. However, as these conditions are uncontrolled the requirement of better algorithms has become essential. We propose a Bidimensional Empirical Mode Decomposition-based unlighting method that preprocesses the luminance and the reflectance parts of an image. First, three luminance components are estimated using Bidimensional Intrinsic Mode Functions residuals. Second, a shadow removal procedure using recursive Retinex is applied. Third, the reflectance part is denoised using mean-Gaussian filters. After that, a new image is created multiplying each shadow-free luminance by the reflectance. The final output is obtained using the geometric mean on the newly acquired images. This algorithm has been tested in two 3D- 2D face recognition databases: UHDB11 and FRGCv2.0. The performance of BEMDU demonstrates an improvement of up to 15.42% when compared with the AELM, LBEMD, PittPatt, the baseline, and EA algorithms.

Original language	English (US)
Title of host publication	2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	19-23
Number of pages	5
ISBN (Electronic)	9781479988822
DOIs	https://doi.org/10.1109/WIFS.2014.7084297
State	Published - 2014
Event	2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014 - Atlanta, United States Duration: Dec 3 2014 → Dec 5 2014

Publication series

Name	2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014

Conference

Conference	2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014
Country/Territory	United States
City	Atlanta
Period	12/3/14 → 12/5/14

Keywords

EMD
Face Recognition
Retinex
Unlighting
illumination preprocessing

ASJC Scopus subject areas

Computer Science Applications
Information Systems
Information Systems and Management
Safety, Risk, Reliability and Quality

Access to Document

10.1109/WIFS.2014.7084297

Cite this

Ochoa-Villegas, M. A., Nolazco-Flores, J. A., Barron-Cano, O., & Kakadiaris, I. A. (2014). Bidimensional empirical mode decomposition-based unlighting for face recognition. In 2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014 (pp. 19-23). Article 7084297 (2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/WIFS.2014.7084297

Bidimensional empirical mode decomposition-based unlighting for face recognition. / Ochoa-Villegas, Miguel A.; Nolazco-Flores, Juan A.; Barron-Cano, Olivia et al.
2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014. Institute of Electrical and Electronics Engineers Inc., 2014. p. 19-23 7084297 (2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Ochoa-Villegas, MA, Nolazco-Flores, JA, Barron-Cano, O & Kakadiaris, IA 2014, Bidimensional empirical mode decomposition-based unlighting for face recognition. in 2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014., 7084297, 2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014, Institute of Electrical and Electronics Engineers Inc., pp. 19-23, 2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014, Atlanta, United States, 12/3/14. https://doi.org/10.1109/WIFS.2014.7084297

Ochoa-Villegas MA, Nolazco-Flores JA, Barron-Cano O, Kakadiaris IA. Bidimensional empirical mode decomposition-based unlighting for face recognition. In 2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014. Institute of Electrical and Electronics Engineers Inc. 2014. p. 19-23. 7084297. (2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014). doi: 10.1109/WIFS.2014.7084297

Ochoa-Villegas, Miguel A. ; Nolazco-Flores, Juan A. ; Barron-Cano, Olivia et al. / Bidimensional empirical mode decomposition-based unlighting for face recognition. 2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014. Institute of Electrical and Electronics Engineers Inc., 2014. pp. 19-23 (2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014).

@inproceedings{6f13950839cf4dc9832476002421b8ee,

title = "Bidimensional empirical mode decomposition-based unlighting for face recognition",

abstract = "A face recognition system must be capable of handling facial data with head pose variations or different illumination conditions. However, as these conditions are uncontrolled the requirement of better algorithms has become essential. We propose a Bidimensional Empirical Mode Decomposition-based unlighting method that preprocesses the luminance and the reflectance parts of an image. First, three luminance components are estimated using Bidimensional Intrinsic Mode Functions residuals. Second, a shadow removal procedure using recursive Retinex is applied. Third, the reflectance part is denoised using mean-Gaussian filters. After that, a new image is created multiplying each shadow-free luminance by the reflectance. The final output is obtained using the geometric mean on the newly acquired images. This algorithm has been tested in two 3D- 2D face recognition databases: UHDB11 and FRGCv2.0. The performance of BEMDU demonstrates an improvement of up to 15.42% when compared with the AELM, LBEMD, PittPatt, the baseline, and EA algorithms.",

keywords = "EMD, Face Recognition, Retinex, Unlighting, illumination preprocessing",

author = "Ochoa-Villegas, {Miguel A.} and Nolazco-Flores, {Juan A.} and Olivia Barron-Cano and Kakadiaris, {Ioannis A.}",

note = "Publisher Copyright: {\textcopyright} 2014 IEEE.; 2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014 ; Conference date: 03-12-2014 Through 05-12-2014",

year = "2014",

doi = "10.1109/WIFS.2014.7084297",

language = "English (US)",

series = "2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "19--23",

booktitle = "2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014",

address = "United States",

}

TY - GEN

T1 - Bidimensional empirical mode decomposition-based unlighting for face recognition

AU - Ochoa-Villegas, Miguel A.

AU - Nolazco-Flores, Juan A.

AU - Barron-Cano, Olivia

AU - Kakadiaris, Ioannis A.

PY - 2014

Y1 - 2014

N2 - A face recognition system must be capable of handling facial data with head pose variations or different illumination conditions. However, as these conditions are uncontrolled the requirement of better algorithms has become essential. We propose a Bidimensional Empirical Mode Decomposition-based unlighting method that preprocesses the luminance and the reflectance parts of an image. First, three luminance components are estimated using Bidimensional Intrinsic Mode Functions residuals. Second, a shadow removal procedure using recursive Retinex is applied. Third, the reflectance part is denoised using mean-Gaussian filters. After that, a new image is created multiplying each shadow-free luminance by the reflectance. The final output is obtained using the geometric mean on the newly acquired images. This algorithm has been tested in two 3D- 2D face recognition databases: UHDB11 and FRGCv2.0. The performance of BEMDU demonstrates an improvement of up to 15.42% when compared with the AELM, LBEMD, PittPatt, the baseline, and EA algorithms.

AB - A face recognition system must be capable of handling facial data with head pose variations or different illumination conditions. However, as these conditions are uncontrolled the requirement of better algorithms has become essential. We propose a Bidimensional Empirical Mode Decomposition-based unlighting method that preprocesses the luminance and the reflectance parts of an image. First, three luminance components are estimated using Bidimensional Intrinsic Mode Functions residuals. Second, a shadow removal procedure using recursive Retinex is applied. Third, the reflectance part is denoised using mean-Gaussian filters. After that, a new image is created multiplying each shadow-free luminance by the reflectance. The final output is obtained using the geometric mean on the newly acquired images. This algorithm has been tested in two 3D- 2D face recognition databases: UHDB11 and FRGCv2.0. The performance of BEMDU demonstrates an improvement of up to 15.42% when compared with the AELM, LBEMD, PittPatt, the baseline, and EA algorithms.

KW - EMD

KW - Face Recognition

KW - Retinex

KW - Unlighting

KW - illumination preprocessing

UR - http://www.scopus.com/inward/record.url?scp=84929259229&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84929259229&partnerID=8YFLogxK

U2 - 10.1109/WIFS.2014.7084297

DO - 10.1109/WIFS.2014.7084297

M3 - Conference contribution

AN - SCOPUS:84929259229

T3 - 2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014

SP - 19

EP - 23

BT - 2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014

Y2 - 3 December 2014 through 5 December 2014

ER -

Bidimensional empirical mode decomposition-based unlighting for face recognition

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this