Vol. 40, Issue 4, pp. 843-861 (2010)

Abstract

The requirement to determine the identity of a person is becoming more and more important in the present information society. There are a number of biometric based identification and verification systems on the market for many civilian, military and forensic applications. Amongst the reliable biometric systems, fingerprint authentication plays a major role, as fingerprints are not just related to criminality anymore. Several automated fingerprint identification or verification systems dictate the market and some apply pores as additional feature to the traditionally applied minutiae systems to improve the system performance. There is, however, a financial trade off between detection of pores and the complexity of the applied optical system. In this paper, the impact of using strike light and the application of a specific wavelength to improve the reflection on the skin is presented to improve the imaging performance of existing automated fingerprint identification or verification systems (AFIS or AFVS). The results are fingerprint images with distinct pores necessary for the image and feature processing. With the introduction of computers it became possible to automate data of fingerprints. The template (original) fingerprint images are compared with the live scan images, resulting in a match score. This match score is defined in terms of the number of false accepts and false reject errors and determines the performance of the system. Despite the inverse relation of these error rates, the lower both rates, the better the system. The final result of the introduced unique features is a system with error rates significantly better than with systems applying standard prism based white light sensors.
The application of all presently available fingerprint sensor systems has been extended significantly after the September 11-th incident.