Multi Modal Biometric Systems: A State of the Art Survey

摘要

In its most general definition, Biometrics refers to the science of automatic recognition of individuals based on some specific physiological and/or behavioural features. A biometric system that is based on one single biometric identifier does not always come to meet the desired performance requirements; multimodal biometric systems come to be an emergent trend. In this paper, we discuss the most commonly used unimodal biometric systems ranging from signature identification, facial recognition, DNA identification, speech authentication, hand geometry recognition, iris recognition and fingerprint identification. However, performances of such unimodal biometric recognition systems can degrade quickly when the input biometric traits suffer extensive variations; the importance of multimodal biometric systems is hence highlighted. Recent advances and application of multimodal biometric systems are then presented.