Enhancing the performance of multimodal biometric authentication system

The term Biometric has the relation with "life measurement" however the tenure is commonly interconnected by means of exclusive physiological personality to recognize an individual. A uni-modal biometric system has a diversity of problems such as noisy data, nonuniversality, spoof attacks and undesirable error rate. These limitations can be solved by deploying multimodal biometric system. Multimodal biometric authentication system makes use of two or more individual modalities, like face, iris, retina and fingerprint etc. Here we have anticipated fingerprint and iris traits at feature level extraction. Pre-processing of images of iris and fingerprint is done and features are extracted from them. Our work composed of two foremost sections: Feature extraction of both traits and fusing them before matching and appliance of an encryption technique to boost the security of the fused template and lastly matching stage to authenticate a person.