Literature review on multimodal biometrics

Abstract

As technological reformation is widen, biometric systems substitute knowledge based and token based recognition systems. Confidential data are accessed by the user after the user is recognized by biometric systems. Efforts have been made to acquire more suitable prototype for recognizing human as multimodal biometrics has more severe concern because of noise in the sample and malfunctioning sensing devices. This paper gives a dual study related to multimodal biometrics, including a literature review of the prior work in authentication and the proposed evaluation approaches. First, we classify few epitome studies considered in last decades to show how this problem has been solved until now. Second, the paper gives a introduction about basic principles of the associated evaluation approaches, and then provide an extended evaluation framework based on the enrollment selection and also statistically convincing measures for evaluating quality metrics.