A Novel Application of the Classifier DECOC Based on Fingerprint Identification

Human fingerprints are rich in details, here called "minutiae". In this paper, a fingerprint recognition system based on a novel application of the classifier DECOC to the minutiae extraction and on an optimised matching algorithm will be presented. The minutiae extraction has been performed from fingerprint skeletons. To identify the different shapes and types of minutiae, a Data-driven Error Correcting Output Coding (DECOC) has been adopted to work as a classifier. The proposed classifier has been applied throughout the fingerprint skeleton to locate various minutiae. Extracted minutiae have been used then as identification marks for an automatic fingerprint matching that is based on distance, direction and type between two minutiae.