Fingerprint Verification Using Local Interest Points and Descriptors

Abstract

A new approach to automatic fingerprint verification based on a general- purpose wide baseline matching methodology is here proposed. The approach is not based on the standard ridge-minutiae-based framework. Instead of detecting and matching the standard structural features, local interest points are detected in the fingerprints, then local descriptors are computed in the neighborhood of these points, and afterwards these descriptors are compared using local and global matching procedures. Then, a final verification is carried out by a Bayes classifier. The methodology is validated using the FVC2004 dataset, where competitive results are obtained.