Design, Acquisition and Collection of Various Multimodal Databases to Increase Accuracy and Level of Security

Abstract

Biometric identification and verification technique is an automated sample detection and recognition system that make use of physiological or behavioral characteristics of individuals. Physical biometric traits which include face, fingerprints, palm prints, ear specimens, eye patterns such as iris and retina. Behavioral biometric traits include gait(behaviour), signature, typing speed/patterns, voice patterns etc. Unimodal biometric systems developed for each of these biometric features are not sufficient for gaining accuracy and security. To overcome this problem, multimodal biometrics are developed that apply fusion strategy of multiple biometric modalities or features from two or more than two sources. When we compare Uni-modal biometric systems with multimodal biometric systems then multimodal systems improves recognition accuracy, security issues and system reliability problems. Various unimodal biometric databases are freely available But researchers are facing problems that there are very few multimodal databases are available developing multi modal biometric. This paper presents, collection of various biometric modalities like face, fingerprint, palm print, ear patterns which are obtained in controlled environment using different sensors to develop multimodal biometric system.