Fuzzy logic method for partial fingerprint recognition

Minutia-based algorithms are the preferred methods for matching fingerprints, because of their temporal performances. Although the correlation techniques are considered to be more reliable, they are much slower due to the fact that a large number of alignments (rotations and displacements) need to be taken into account. On the other hand, minutiae-based methods perform poorly when applied to low quality images, and might be impossible to employ for partial fingerprint recognition. This paper proposes a novel fuzzy logic method for matching partial fingerprints. The inputs of the proposed fuzzy logic algorithm are the correlation degree and the relative surface of the input fingerprint.