Multimodal Biometric System: An Overview

Abstract: Information Security is the process by which an organization protects and secures data communications and storage. Biometrics is a field of science which deals with the positive identification of unique individuals based on their physiological or behavioural characteristics. Each biometric has its strength and weakness. No single biometric is expected to effectively meet the requirements of all applications. Biometric products provide improved security over traditional electronic access controls. While some technologies have gained more acceptance then others, it is beyond doubt that the field of access control biometrics has gained a measure of acceptance. In certain situations, the user might find one form of biometric identification is not exact enough for identification. Unimodal biometric system have to contend with a variety of problems such as noisy data, spoof attacks, non-universality, intra-class variations etc. These limitations can be overcome by Multimodal biometric system. Multimodal biometric technology uses more then one biometric identifier to compare the identity of the person for better security. This article describes about Multimodal Biometric technologies, modules and levels of fusions, design issues and modes of operations etc.

Keywords: Fusion, PCA, LDA, Patterns, Minutia, Feature extraction