Secured Multi Modal Biometric System : A Review

Abstract

Multi biometrics is a significant and interesting research area. It is used in identifying individuals for security purposes and increasing security levels. Multi modal biometric provides solution over unimodal biometric system. Various approaches and methods have been studied to improve the accuracy of identification. A state-of-the-art survey on Unimodal biometric limitations, need of multi modal biometrics, conventional methods for identification and different fusion levels are discussed. The integration of machine learning in biometrics is highlighted to improve the accuracy in identification process. The biometric features taken firstly are not similar when they are taken twice. Due to these features, usage of machine learning techniques like Neural Networks, fuzzy logic, evolutionary computing etc. has developed a great demand.