Fingerprint indexing with bad quality areas

摘要

Fingerprint indexing is a key technique in fingerprint identification systems. This strategy allows us to reduce the search space and the occurrences of false acceptance in databases with great size. This paper presents a new triplet based indexing algorithm which uses a new fingerprint representation, based on minutia triplets. This representation is an extension of the triangle set obtained from Delaunay triangulation. Also, a strategy is proposed in order to dismiss bad quality triplets that could affect the accuracy of the indexing process. This proposal shows a good accuracy, even when the fingerprints have bad quality areas.