Fingerprint recognition based on spectral feature extraction

With the advancements in computational techniques and computer technology, biometric-based authentication has a strong potential to be widely utilized in a variety of applications. Among various biometrics, fingerprint-based identification is the most mature and widely accepted technique. In this paper, a fast procedure that exploits the spectral features of the fingerprint to obtain a compact descriptor representation, that is both rotation and translation-invariant, is presented. Experimental results have shown high matching accuracies that are equivalent or even better than those reported in the literature.

[1]  Anil K. Jain,et al.  Fingerprint classification and matching using a filterbank , 2001 .

[2]  Anil K. Jain,et al.  Handbook of Fingerprint Recognition , 2005, Springer Professional Computing.

[3]  Louis Coetzee,et al.  Fingerprint recognition in low quality images , 1993, Pattern Recognit..

[4]  A. J. Willis,et al.  A cost-effective fingerprint recognition system for use with low-quality prints and damaged fingertips , 2001, Pattern Recognit..

[5]  Teddy Ko,et al.  Fingerprint enhancement by spectral analysis techniques , 2002, Applied Imagery Pattern Recognition Workshop, 2002. Proceedings..

[6]  D M Berfanger,et al.  All-digital ring-wedge detector applied to fingerprint recognition. , 1999, Applied optics.

[7]  Anil K. Jain,et al.  FVC2000: Fingerprint Verification Competition , 2002, IEEE Trans. Pattern Anal. Mach. Intell..

[8]  Pauli Kuosmanen,et al.  Wavelet domain features for fingerprint recognition , 2001 .