Nonlinear component analysis as a kernel eigenvalue problem
Neural Computation
Using ear biometrics for personal recognition
IWBRS'05 Proceedings of the 2005 international conference on Advances in Biometric Person Authentication
| Hi-index | 0.00 |
In this paper, a simple multimodal biometrics recognition system having three modalities i.e. Gait, Ear and Face is used and for different biometric traits features Kernel Principal Component method is used for fusion. Because of these biometric traits, our proposed method requires no significant user co-operation and also work from a long distance. The method has been successfully tested on 300 images corresponding to 30 subjects from three databases including ORL face database, USTB ear database and CASIA gait database. The experimental results exhibit excellent recognition performance than Gait, Ear and Face unimodal biometric recognition. As, the every database contain the data of different persons so we can use them only for testing for the given subject.