Multi Biometric System using Efficient Method of DWT

Abstract

A multimodal biometric system is the fusion of one or more biometrics traits. It gives better recognition performance compare to system based on a single biometric modality. This paper proposes a multimodal biometric system using two traits i.e. face and palm print for person recognition. In this proposed method use an efficient technique for an authentication with an average half face and selected window size palm print features. Integrating the features increases robustness of the person authentication. Discrete Wavelet Transform (DWT) is used for feature extraction and Support Vector Machine (SVM) is proposed for classification. The final decision is made by fusion at matching score level. The proposed system consists of three phases: (i)Face Recognition,(ii)Palm Print Recognition and (iii)Multimodal biometrics Fusion method. The proposed method was tested using chimerical datasets consists of the cropped extended Yale database, where the images vary in illumination and expression, IIT palm print database where the images has been acquired using a simple and touch less imaging setup. High recognition performance has been obtained by fusion of these features. The experimental results showed the effectiveness of system in terms of GAR. With Average Half Face only 91.05%, with selected palm print only 90.04% and with fusion of face and palm print 99.47%.The proposed system shows a high degree of success in identifying the individual with reduced computation time and memory storage by using average half face and selected window size palm print features.