Please use this identifier to cite or link to this item: http://hdl.handle.net/10397/14226
Title: Bimodal biometrics based on a representation and recognition approach
Authors: Xu, Y
Zhong, A
Yang, J
Zhang, D 
Keywords: Biometrics
Computer vision
Face recognition
Pattern recognition
Issue Date: 2011
Publisher: SPIE-International Society for Optical Engineering
Source: Optical engineering, 2011, v. 50, no. 3, 037202 How to cite?
Journal: Optical engineering 
Abstract: It has been demonstrated that multibiometrics can produce higher accuracy than single biometrics. This is mainly because the use of multiple biometric traits of the subject enables more information to be used for identification or verification. In this paper, we focus on bimodal biometrics and propose a novel representation and recognition approach to bimodal biometrics. This approach first denotes the biometric trait sample by a complex vector. Then, it represents the test sample through the training samples and classifies the test sample as follows: let the test sample be expressed as a linear combination of all the training samples each being a complex vector. The proposed approach obtains the solution by solving a linear system. After evaluating the effect, in representing the test sample of each class, the approach classifies the test sample into the class that makes the greatest effect. The approach proposed is not only novel but also simple and computationally efficient. A large number of experiments show that our method can obtain promising results.
URI: http://hdl.handle.net/10397/14226
ISSN: 0091-3286
EISSN: 1560-2303
DOI: 10.1117/1.3554740
Appears in Collections:Journal/Magazine Article

Access
View full-text via PolyU eLinks SFX Query
Show full item record

SCOPUSTM   
Citations

44
Last Week
0
Last month
0
Citations as of Oct 19, 2017

WEB OF SCIENCETM
Citations

25
Last Week
0
Last month
0
Citations as of Oct 15, 2017

Page view(s)

53
Last Week
1
Last month
Checked on Oct 15, 2017

Google ScholarTM

Check

Altmetric



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.