Enhancement of Security through Fused Multimodal Biometric System

摘要

Now a days, Person identity is a major issue for security in many applications for example access control, electronic commerce, banking, communications etc. Multimodal biometric technology is the best technology for identification of person than the conventional password based technique. When two or more biometrics used for identification of person then the system called as Multimodal Biometric system. In this paper, we are going to propose a robust security system for personal identification using multimodal biometric system. Three modalities are used in this system face, fingerprint and palm vein. Feature extraction of each modality is done with different algorithm. Local Binary Pattern (LBP) algorithm is used for face feature extraction. Discrete Wavelet transform (DWT) is used for Palm vein feature extraction. And minutiae algorithm is used for finger print feature extraction. Then feature level fusion is done. The proposed method is carried on self generated database images. Finally the more accurate results are achieved. We can further enhance the security adding some encrypted key after fusion.