A new multimodal biometric system using tripled chaotic watermarking approach

Abstract

Recent years, the conventional unimodal biometric system raises up several critical issues of security and integrity of the biometrics data. Consequently, a novel multimodal biometric system which provides better security is proposed in this paper. Fingerprint and face, which form a good combination for a multimodal biometric system is used. The facial eigen-coefficients are first extracted from Principal Component Analysis (PCA), a novel tripled chaotic encryption is then applied to encrypt the eigen-coefficients. Three chaotic maps are involved to enhance system complexity while provides large enough key space to resist from brute-force attacks. The encrypted eigen-coefficients are then embedded into fingerprint using Wavelet Packet Transform (WPT) watermarking scheme. A watermarked fingerprint is created and stored into the smart card. The proposed multimodal biometric system possesses the features of multimodal biometric, encryption and watermarking. The work can be applied to the smart card and credit card application. The simulation results show that the proposed system has a good performance.