Exploring fingerprint matching through a topology-based perspective

In the recent decade, the rapid development of biometric technologies for security applications spurred increased interest in both academic research community and industrial sectors. Still, the fingerprint recognition remains one of the core methodologies for person identification. Known since the time of the Babylonians, they used their thumbprints on clay tablets to verify their identity around 2,000 B.C. While the approach remains a reliable, efficient and a commonly accepted biometric, the methodology develops rapidly. This paper presents a survey of topology-based techniques for fingerprint recognition, and describes an original fingerprint alignment and matching scheme based on two powerful computational geometry data structures: the Voronoi diagram and the Delaunay triangulation.