Image mosaicing algorithm for rolled fingerprint construction

Fingerprint identification is one of the most important biometric authentication methods. However, current devices for recording digital fingerprints can only capture plain-touch fingerprints. Rolled fingerprints have much more information for recognition, so a method is needed to construct a rolled fingerprint from a series of plain-touch fingerprints. This paper presents a novel algorithm for image mosaicing for real time rolled fingerprint construction in which the images are assembled with corrections to create a smooth, non-fragmented rolled fingerprint in real time. Experimental results demonstrate its effectiveness by comparing it with other conventional algorithms.