M.S. ALTARAWNEH, W.L. WOO, and S.S. DLAY

Abstract

 Fingerprint based system reliability is critical for image validity and quality. As, good quality images make the system performance more robust. However, it is always very difficult to obtain good quality images in practice and image quality degradation leads to invalid images. A hybrid method for fingerprint image validity calculation is proposed that combines validity check, using statistical weight calculation and Gabor spectrum approach for image quality assessment. The proposed scheme is investigated using a manual validity check survey and is compared with NIST fingerprint image for image quality. Experimental results on FVC2004 and TIMA database show that the proposed method significantly improves false acceptance and true acceptance rates.

  The 7th WSEAS International Conference on SIGNAL PROCESSING, ROBOTICS and AUTOMATION (ISPRA '08), Cambridge, UK, February 20-22, 2008

http://portal.acm.org/citation.cfm?id=1416152