An Automated Fingerprint Identification System Study Based On Multi-Patterns

Fingerprint-based identifications (eg:AFIS), though are the most common use in biometrics fields, still now have difficult problems in fingerprint identification's core technologies. Particularly of the fingerprint which is poor or damaged. With the development of electronic computer and information technology, the identification technology using biometric have drawn more attention.The paper describes the following algorithms based on the predecessor's fruit:First is the pretreatments on the gathered fingerprints, the fingerprint pretreatments are refers to the fingerprint regularization, the enhancement, the binarization which are a very important intermediate step in the extraction of minutiaes. We use an effective way to enhance the poor quality fingerprints based on Gabor filter. The following algorithms include the fingerprints binarization and the fingerprints thinning. The minutiaes extraction mainly uses 8-neighborhood of points after the fingerprint binarization. During the matching this paper uses the algorithm with the Minutiaes Matching method and the Structure Matching method. The addition information we use to the minutiaes are evidently simplified the matching of the fingerprints according to the experimental result.