Fingerprint Matching on Small Images

The uniqueness of the ridges’ flow pattern is the basis of forensic applications on fingerprints, the accuracy of matching two or more fingerprints being its most important issue. A big problem arises when the fingerprint image used is small. In this paper we aim at introducing a method of analyzing a small image so that valid minutiae points can be extracted. For this purpose a constraint satisfaction based technique for finding a match between a candidate fingerprint with one from a database of fingerprints is introduced. We also provide experimental results and a comparison with other similar existing approaches.