Fingerprint matching using local alignment based on multiple pairs of reference minutiae

Guo Cao, Yuan Mei, Zhihong Mao, Quan-Sen Sun

Abstract: A novel minutiae-based method is proposed to match deformed fingerprints. We combine minutiae structures and descriptors to obtain multiple pairs of reference minutiae and globally align two sets of minutiae to get a common overlapping region based on these pairs of reference minutiae. At the matching stage, our proposed approach aligns local areas in two fingerprints with new reference minutiae pairs that are selected from matched minutiae pairs. After registration of the fingerprints according to the local correspondence, the number of matched minutiae can be counted using tight thresholds. Experimental results confirm that the proposed algorithm is effective for fingerprint matching with nonlinear distortions.