Cluster Approach for Fingerprint Identification

摘要

The paper describes a clustering method to perform one-to-many fingerprint comparisons. The algorithm starts with finding the minutiae of a series of reference fingerprints. Then, it tries to group the minutiae into clusters. Each of them is made of those minutiae that have similar features. The enrolment phase consists in acquiring several images of the same finger and finding the common clusters. This is meant to be done for many images of different fingers. When we have to verify the matching with a new fingerprint, we just extract its minutiae and compare them with the clusters of several groups of images. The experiment results demonstrate that this similarity-searching approach proves suitable for one-to-many matching of fingerprints on large-scale databases. With the feedback module the proposed fingerprint identification scheme has inspiring identification performance of application.