Ear and fingerprint biometrics for personal identification

Multimodal biometrics provides ultra-secure authentication using multiple biometric traits. In this paper, ear and fingerprint biometrics are used for the personal identification. A novel technique of edge interaction point detection (EIPD) is used to determine the ear features. Fingerprint features are identified by line based connected component analysis and its feature vectors are generated using EIPD system. Fusion at the feature level is carried out through concatenation for features. Neural network based back propagation network is used for an identity verification system. As a first level of verification, matching is performed using ear biometrics alone. At this level, with horizontal matrix method 88% GAR and with the combination of horizontal and vertical matrix 92% GAR is achieved. In second level of verification, feature vector of fused matrices of ear and fingerprint is used and 100% GAR is obtained.