Liveness Detection in Biometrics

The biometric systems, oriented in this chapter especially on fingerprints, have been introduced in the previous chapters. The functionality of such systems is influenced not only by the used technology, but also by the surrounding environment (including skin or other diseases). Biased or damaged biometric samples could be rejected after revealing their poor quality, or may be enhanced, what leads to the situation that samples, which would be normally rejected, are accepted after the enhancement process. But this process could present also a risk, because the poor quality of a sample could be caused not only by the sensor technology or the environment, but also by using an artificial biometric attribute (imitation of a finger(print)). Such risk is not limited just to the deceptional technique, but if we are not able to recognize whether an acquired biometric sample originates from a genuine living user or an impostor, we would then scan an artificial fake and try to enhance its quality using an enhancement algorithm. After a successful completion of such enhancement, such fake fingerprint would be compared with a template and if a match is found, the user is accepted, notwithstanding the fact that he can be an impostor! Therefore the need of careful liveness detection, i.e. the recognition whether an acquired biometric sample comes from a genuine living user or not, is crucial.