An Improved Approach in Core Point Detection Algorithm for Fingerprint Recognition

Abstract

Core Point Detection is used when comparison of two fingerprint based on alignment is done. To extract the core point is still open issue in fingerprint recognition. The success rate of fingerprint recognition as well as to speed up the process core point is used. The proposed core point detection algorithm follows four steps: normalization, ridge orientation estimation, smoothing and core point detection. To test the experimental result and performance of proposed algorithm carried out using FVC2000 and FingerDOS databases and prove that, the FMR and FNMR for FVC2000 is 2 and 1.2 and for FingerDOS is 0 and 1.67 respectively. The accuracy of FVC2000 is 98.4% and FingerDOS is 99.16%. The average execution time of FVC2000 is less than 0.72 and for FingerDOS is 0.25 second.