Noise robust multimodal biometric person authentication system using face, speech and signature features

In this work, we present a multimodal biometric system using face, speech and signature features which is robust to noise. Face recognition is done using subspace, principal component analysis (PCA) and linear discriminant analysis (LDA) techniques. Speaker recognition system is built using mel frequency cepstral coefficients (MFCC) for feature extraction and vector quantization (VQ) for pattern matching. An off-line signature recognition system is built using vertical and horizontal projection profiles (VPP, HPP) and discrete cosine transform (DCT) for feature extraction. A multimodal biometric database with face, speech and signature biometric features has been collected for 30 users. A multimodal biometric system is built using score level fusion. Sum rule was used for the fusion of the biometric scores. Experimental results show the efficacy of the multimodal biometric system when the biometric data is affected by noise.