(For USM Staff/Student Only)

EngLib USM > Ω School of Electrical & Electronic Engineering >

Fusion of vein and geometry for finger based biometric system

Fusion of vein and geometry for finger based biometric system / Seng Lee Jun
Sistem biometrik adalah satu daripada pengesahan identiti popular yang digunakan untuk pelbagai tujuan kawalan akses identiti. Pengesahan salur darah jari merupakan satu daripada sistem biometrik ekaragam yang tersedia dalam bidang biometrik. Walaupun terdapat pelbagai kelebihan yang ditawarkan oleh pengesahan salur darah jari, beberapa kelemahan yang dihadapi oleh pengesahan salur darah jari telah mengehadkan prestasi dan penerimaannya untuk penggunaan kawalan akses yang sulit. Oleh itu, satu sistem biometrik berbilang-ragam telah dicadangkan dengan menggabungkan pengesahan geometri jari dengan pengesahan salur darah jari. Sistem biometrik berbilang-ragam mampu meningkatkan prestasi sistem biometrik yang berasaskan jari sampai di tahap ketepatan dan kebolehpercayaan yang lebih tinggi. Pelaksanaan sistem biometrik berbilang-ragam dalam projek ini melibatkan tiga peringkat: pelaksanaan pengesahan geometri jari, pelaksanaan pengesahan salur darah jari dan gabungan pengesahan salur darah jari dan pengesahan geometri jari sebagai sistem biometrik berbilang-ragam. Dalam pelaksanaan pengesahan geometri jari, ketebalan jari digunakan sebagai ciri biometrik. Prestasi pengesahan geometri adalah di bawah jangkaan, di mana kadar ralat sama ialah 40.7024%. Manakala dalam pelaksanaan pengesahan salur darah jari, band fasa terhad sekaitan imej salur darah jari telah digunakan. Prestasi pengesahan salur darah jari adalah sangat baik, di mana kadar ralat sama adalah serendah 2.2189%. Dalam peringkat terakhir, jumlah wajaran digunakan untuk menggabungkan skor sepadan pengesahan geometri jari dan pengesahan salur darah jari. Kadar ralat sama gabungan pengesahan geometri jari dan pengesahan salur darah jari ialah 10.5668%. Dari keputusan yang diperolehi, ia ditunjukkan bahawa gabungan pengesahan salur darah jari dan pengesahan geometri jari ialah satu pendekatan berpotensi untuk meningkatkan lagi prestasi sistem biometrik. _______________________________________________________________________________________________________ Biometric system is one of the popular identity authentication used for various identity access control purpose. Finger vein authentication is one of the popular unimodal biometric system available in the field of biometrics. Despite there are various advantages offered by finger vein authentication, several inabilities faced by finger vein biometrics had limited its performance and adoption for highly confidential access control application. Hence, a new multimodal biometric system is proposed by fusing finger geometry authentication with finger vein authentication. The proposed multimodal biometric system is able to improve the performance of finger based biometric system to a higher level of accuracy, preciseness and reliability. Implementation of multimodal biometric system in this project involves three stages: implementation of finger geometry biometric system, implementation of finger vein biometric system and score level fusion of finger geometry and finger vein biometric system. In implementation of finger geometry biometric system, finger thickness is used as the biometric feature to be used for matching. The performance of geometry biometric system is under expectation, where the equal error rate (EER) is 40.7024%. While in implementation of finger vein biometric system, Band Limit Phase Only Correlation of the finger vein image is used for matching. The performance of vein biometric system is remarkable, where the EER is as low as 2.2189%. In the last stage, weighted SUM is adopted to fuse the matching scores of finger geometry biometric system and finger vein biometric system. The EER of the fusion of geometry biometric system and vein biometric system is 10.5668%. From the results, it is shown that fusion of finger vein recognition and finger geometry recognition is a potential approach to enhance the performance of biometric system.
Contributor(s):
Seng Lee Jun - Author
Primary Item Type:
Final Year Project
Identifiers:
Accession Number : 875004756
Language:
English
Subject Keywords:
Biometric; authentication; vein
First presented to the public:
6/1/2012
Original Publication Date:
7/2/2020
Previously Published By:
Universiti Sains Malaysia
Place Of Publication:
School of Electrical & Electronic Engineering
Citation:
Extents:
Number of Pages - 96
License Grantor / Date Granted:
  / ( View License )
Date Deposited
2020-07-02 14:58:35.381
Submitter:
Mohd Jasnizam Mohd Salleh

All Versions

Thumbnail Name Version Created Date
Fusion of vein and geometry for finger based biometric system1 2020-07-02 14:58:35.381