Accurate localizations of reference points in a fingerprint image
PReMI'11 Proceedings of the 4th international conference on Pattern recognition and machine intelligence
Feature-level fusion of fingerprint and finger-vein for personal identification
Pattern Recognition Letters
| Hi-index | 0.00 |
Fingerprint recognition is a widely used biometricidentification mechanism. In case of correlation based fingerprintrecognition detection of a consistent registration point is a crucialissue; this point can be a core point of a fingerprint. Manytechniques have been proposed but success rate is highlydependent on input used and accurate core point detection is stillan open issue. Here we discuss a core point detection algorithmwhich is computationally simple and gives consistent detection ofa core point. The proposed technique is based on orientation fieldof the fingerprint which is calculated using gradient basedtechnique and optimized neighborhood averaging to generate asmoother orientation field, on which we operate a speciallydesigned mask for detecting core point as the orientation field inthe region of the core point is different than the other area.Though all fingerprints don’t possess core point still thisalgorithm is useful to detect high curvature regions. Thisalgorithm is helpful in the development of correlation basedAutomatic Fingerprint Recognition System (AFIS).