Fingerprint Matching Using Global Minutiae and Invariant Moments

In this paper, a fingerprint matching algorithm based on global minutiae and invariant moments is proposed. Although minutiae features based method is popular in most case of fingerprint matching, minutiae are difficult to be extracted robustly in low quality image and easy to deduce false recognition. Combination of other discriminatory features can effectively strengthen the performance of the matching. A fingerprint image is first preprocessed for minutiae and reference point determination. Then, the reference point is used to align the template fingerprint and input fingerprint. Based on the aligned fingerprints, combined features are constructed from all the minutiae and their invariant moments, and a global matching algorithm is proposed to match the template and input fingerprint. The experimental results show higher performance comparing with a public algorithm over FVC2004 databases.