An efficient technique of multimodal biometrics using fusion of face and iris features

Multimodal biometrics system increase robustness by removing a dependency on one single biometric traits. In recent years, Multimodal biometrics has received large attention in research. Multimodal Biometrics are more secure than unimodal biometric as even if one biometric trait fail other is present. In this paper, we used a combination of face and iris biometric traits. We used Viola Jones algorithms for detection of face in an image and circular Hough transform to segmented iris part in the image. For feature extraction, we used Modular Principal Component analysis (MPCA) algorithm for face and 2D Gabor filter for iris. The obtained features of face and iris biometric traits are combined using Znorm fusion technique, and forming a single vector. This paper proposes unique own generated database of used modalities.