Rotation invariant iris feature extraction using Gaussian Markov random fields with non-separable wavelet

Jing Huang, Xinge You, Yuan Yuan, Feng Yang, Lin Lin

Research output: Contribution to journalArticle

Abstract

Rotation invariance is important for an iris recognition system since changes of head orientation and binocular vergence may cause eye rotation. The conventional methods of iris recognition cannot achieve true rotation invariance. They only achieve approximate rotation invariance by rotating the feature vector before matching or unwrapping the iris ring at different initial angles. In these methods, the complexity of the method is increased, and when the rotation scale is beyond the certain scope, the error rates of these methods may substantially increase. In order to solve this problem, a new rotation invariant approach for iris feature extraction based on the non-separable wavelet is proposed in this paper. Firstly, a bank of non-separable orthogonal wavelet filters is used to capture characteristics of the iris. Secondly, a method of Markov random fields is used to capture rotation invariant iris feature. Finally, two-class kernel Fisher classifiers are adopted for classification. Experimental results on public iris databases show that the proposed approach has a low error rate and achieves true rotation invariance.

Original languageEnglish
Pages (from-to)883-894
Number of pages12
JournalNeurocomputing
Volume73
Issue number4-6
Early online date18 Nov 2009
DOIs
Publication statusPublished - Jan 2010

Fingerprint

Iris
Feature extraction
Invariance
Binoculars
Classifiers
Head
Databases

Bibliographical note

Bayesian Networks / Design and Application of Neural Networks and Intelligent Learning Systems (KES 2008 / Bio-inspired Computing: Theories and Applications (BIC-TA 2007)

Keywords

  • Iris recognition
  • Kernel Fisher classifiers
  • Markov random fields
  • Non-separable wavelet transform

Cite this

Huang, Jing ; You, Xinge ; Yuan, Yuan ; Yang, Feng ; Lin, Lin. / Rotation invariant iris feature extraction using Gaussian Markov random fields with non-separable wavelet. In: Neurocomputing. 2010 ; Vol. 73, No. 4-6. pp. 883-894.
@article{58d5f927667a4da79b4e32a513d24933,
title = "Rotation invariant iris feature extraction using Gaussian Markov random fields with non-separable wavelet",
abstract = "Rotation invariance is important for an iris recognition system since changes of head orientation and binocular vergence may cause eye rotation. The conventional methods of iris recognition cannot achieve true rotation invariance. They only achieve approximate rotation invariance by rotating the feature vector before matching or unwrapping the iris ring at different initial angles. In these methods, the complexity of the method is increased, and when the rotation scale is beyond the certain scope, the error rates of these methods may substantially increase. In order to solve this problem, a new rotation invariant approach for iris feature extraction based on the non-separable wavelet is proposed in this paper. Firstly, a bank of non-separable orthogonal wavelet filters is used to capture characteristics of the iris. Secondly, a method of Markov random fields is used to capture rotation invariant iris feature. Finally, two-class kernel Fisher classifiers are adopted for classification. Experimental results on public iris databases show that the proposed approach has a low error rate and achieves true rotation invariance.",
keywords = "Iris recognition, Kernel Fisher classifiers, Markov random fields, Non-separable wavelet transform",
author = "Jing Huang and Xinge You and Yuan Yuan and Feng Yang and Lin Lin",
note = "Bayesian Networks / Design and Application of Neural Networks and Intelligent Learning Systems (KES 2008 / Bio-inspired Computing: Theories and Applications (BIC-TA 2007)",
year = "2010",
month = "1",
doi = "10.1016/j.neucom.2009.09.016",
language = "English",
volume = "73",
pages = "883--894",
journal = "Neurocomputing",
issn = "0925-2312",
publisher = "Elsevier",
number = "4-6",

}

Rotation invariant iris feature extraction using Gaussian Markov random fields with non-separable wavelet. / Huang, Jing; You, Xinge; Yuan, Yuan; Yang, Feng; Lin, Lin.

In: Neurocomputing, Vol. 73, No. 4-6, 01.2010, p. 883-894.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Rotation invariant iris feature extraction using Gaussian Markov random fields with non-separable wavelet

AU - Huang, Jing

AU - You, Xinge

AU - Yuan, Yuan

AU - Yang, Feng

AU - Lin, Lin

N1 - Bayesian Networks / Design and Application of Neural Networks and Intelligent Learning Systems (KES 2008 / Bio-inspired Computing: Theories and Applications (BIC-TA 2007)

PY - 2010/1

Y1 - 2010/1

N2 - Rotation invariance is important for an iris recognition system since changes of head orientation and binocular vergence may cause eye rotation. The conventional methods of iris recognition cannot achieve true rotation invariance. They only achieve approximate rotation invariance by rotating the feature vector before matching or unwrapping the iris ring at different initial angles. In these methods, the complexity of the method is increased, and when the rotation scale is beyond the certain scope, the error rates of these methods may substantially increase. In order to solve this problem, a new rotation invariant approach for iris feature extraction based on the non-separable wavelet is proposed in this paper. Firstly, a bank of non-separable orthogonal wavelet filters is used to capture characteristics of the iris. Secondly, a method of Markov random fields is used to capture rotation invariant iris feature. Finally, two-class kernel Fisher classifiers are adopted for classification. Experimental results on public iris databases show that the proposed approach has a low error rate and achieves true rotation invariance.

AB - Rotation invariance is important for an iris recognition system since changes of head orientation and binocular vergence may cause eye rotation. The conventional methods of iris recognition cannot achieve true rotation invariance. They only achieve approximate rotation invariance by rotating the feature vector before matching or unwrapping the iris ring at different initial angles. In these methods, the complexity of the method is increased, and when the rotation scale is beyond the certain scope, the error rates of these methods may substantially increase. In order to solve this problem, a new rotation invariant approach for iris feature extraction based on the non-separable wavelet is proposed in this paper. Firstly, a bank of non-separable orthogonal wavelet filters is used to capture characteristics of the iris. Secondly, a method of Markov random fields is used to capture rotation invariant iris feature. Finally, two-class kernel Fisher classifiers are adopted for classification. Experimental results on public iris databases show that the proposed approach has a low error rate and achieves true rotation invariance.

KW - Iris recognition

KW - Kernel Fisher classifiers

KW - Markov random fields

KW - Non-separable wavelet transform

UR - http://www.scopus.com/inward/record.url?scp=75749093884&partnerID=8YFLogxK

U2 - 10.1016/j.neucom.2009.09.016

DO - 10.1016/j.neucom.2009.09.016

M3 - Article

VL - 73

SP - 883

EP - 894

JO - Neurocomputing

JF - Neurocomputing

SN - 0925-2312

IS - 4-6

ER -