Sunil Kumar Singla and Ajat Shatru Arora

pp. 247 - 254

Abstract

Fingerprint is one of the most promising method among all the biometric techniques and has been used for the
personal authentication for a long time because of its wide acceptance and reliability. Features (Minutiae) are extracted from
the fingerprint in question and are compared with the features already stored in the database for authentication. Crossing
number (CN) is the most commonly used minutiae extraction method for fingerprints. In this paper, a new Karnaugh-Map
based fingerprint minutiae extraction method has been proposed and discussed. In the proposed algorithm the 8 neighbors
of a pixel in a 3×3 window are arranged as 8 bits of a byte and corresponding hexadecimal (hex) value is calculated. These
hex values are simplified using standard Karnaugh-Map (K-map) technique to obtain the minimized logical expression.
Experiments conducted on the FVC2002/Db1_a database reveals that the developed method is better than the crossing
number (CN) method.