A new approach to random walk matching of fingerprint recognition

Author

Lee, Ho Young ; Yu, Kyung Deok ; Choi, Tae Young

Author_Institution

Sch. of Electron. Eng., Ajou Univ., Suwon, South Korea

fYear

2004

fDate

18-19 Nov. 2004

Firstpage

39

Lastpage

42

Abstract

This paper presents a new algorithm called the random walk matching for fingerprint verification (or identification), which is based on the distribution of minutia points. First, it sorts minutia pairs according to their distances. Then, it makes a minutia tree based on the coordinate and the angular distance. Unlike existing algorithms, the algorithm shows robustness with respect to the orientation and the type of minutia. To test its efficiency, it is tested with the fingerprint images of FVC2002 database. The result shows an EER of 8.84%.

Keywords

feature extraction; fingerprint identification; police data processing; EER; FVC2002 database fingerprint images; algorithm robustness; angular distance; fingerprint identification; fingerprint recognition; minutia coordinate; minutia orientation; minutia pair distances; minutia points distribution; minutia tree; random walk matching; Algorithm design and analysis; Bifurcation; Fingerprint recognition; Fingers; Image databases; Image matching; Pattern matching; Robustness; Testing; Transforms;

fLanguage

English

Publisher

ieee

Conference_Titel

Intelligent Signal Processing and Communication Systems, 2004. ISPACS 2004. Proceedings of 2004 International Symposium on

Print_ISBN

0-7803-8639-6

Type

conf

DOI

10.1109/ISPACS.2004.1439011

Filename

1439011