Partial Fingerprint Alignment and Matching Through Region-Based Approach

Abstract

One of the most important issues in fingerprint recognition is that the impressions of different fingers may appear very similar to each other (high inter-class similarity) and likewise, different acquisition of the same finger may be dissimilar from each other (high intra-class variation). To compensate for high intra-class variation and inter-class similarity, this paper addresses two challenging issues in partial fingerprint recognition which are alignment and computing the similarity score. Rotational and displacement variation between registered and query fingerprints occur during scanning. As a consequence, fingerprint alignment is essential to correct for the rotational and translation difference between fingerprints prior to matching. By perfectly aligning the fingerprints, the matching could be simplified. Considering the possibility of high intra-class variation and inter-class similarity, how the similarity score of two fingerprints is computed plays an important role in system's match or non-match decision.