Fingerprints matching based on line similarity

A novel fingerprint matching algorithm based on the matching of lines is introduced in this paper. The algorithm were designed in three steps. In the first phase, an inaccurate set of matched line pairs were built by matching lines between two fingerprints. In the second phase, the spurious pairs were removed from the set by evaluating the similarity of triangles composed by line and central point. In the third phase, pairs of minutiae that composed the line pair were used to measure the comprehensive similarity of two fingerprints. Experiments show the proposed method works well over the fingerprint databases of FVC2002.