Polarization-enhanced fingerprint verification system using fringe-adjusted joint-transform correlator

Biometrics has an important role in today's security and identity verification process. Fingerprint identification is the most reliable technique used in biometric systems. However, manual fingerprint systems are being replaced with automatic fingerprint identification systems. Fingerprint verification systems have a very important role in many civilian and military applications. So, the reliability, accuracy and integrity of the systems should be high. There has been renewed interest in the development of efficient systems, which has fueled a great amount of research in building this kind of systems. In this research a novel polarization-enhanced fingerprint identification/verification technique is proposed. An image acquisition system was built to enroll the fingerprints. Stokes vector imagery is used to explore various polarization states of the input scene. These polarization-enhanced images are used in a correlation-based algorithm for verification and identification. The algorithm is based on fringe-adjusted joint-transform correlation principle (FJTC). This technique yields significantly better correlation output than classical or binary joint transform-correlation (JTC). Different polarization states are studied by applying the images to FJTC. The performance metrics of each case are studied and compared. Finally the effect of distortion on the fingerprint is also analyzed.