Fingerprint alignment using a two stage optimization

This paper presents a novel approach to fingerprint alignment based on the optimization of cost functions. The optimization is performed in two stages: the first stage provides a robust initial registration based on non-minutiae features, and the second stage proceeds by fine tuning the alignment parameters using minutiae. This approach represents a significant departure from traditional fingerprint matching algorithms that rely heavily on minutiae features for both registration and verification. The resulting algorithm is not only simple and intuitive, but is also robust, efficient, and accurate. Several alternative alignment algorithms have been implemented, and their results are compared using an FVC2002 dataset. An EER of 1.6% has been achieved for the proposed algorithm.