Fingerprint Alignment Using Line Segments

In this paper, it is provided statistical evidence that supports fingerprint minutiae matching algorithms which use line segments formed by pairs of minutiae as candidates for pivots. This pivots are used to superimpose the analyzed fingerprint templates. Also in this work an algorithm to improve the matcher performance for uncontrolled acquisition systems is proposed. This algorithm employs a technique to sort the minutiae list in templates increasing the chances that corresponding line segments in two templates are tested in the early algorithm iterations. The analysis and the proposed algorithm are validated with data from FVC2000 and FVC2002 databases.