K-Nearest Neighbor Classification Approach for Face and Fingerprint at Feature Level Fusion

Biometric system that based on single biometric called uni-modal biometrics usually suffers from problems like imposter's attack or hacking, unacceptable error rate and low performance. So the need of using multimodal biometric system arises in such cases. The aim of this paper is to study the fusion at feature extraction level for face and fingerprint. The proposed system fuses the two traits at feature extraction level by first making the feature sets compatible for concatenation and then reducing the feature sets to handle the "problem of curse of dimensionality". After concatenation these features are classified. Features of both modalities are extracted using Gabor filter and Principal Component Analysis (PCA). K-Nearest Neighbour classifier is used to classify the different people in the database. The experimental results reveal that the fusion of more than one biometric trait at feature level fusion with the K-Nearest Neighbor technique exhibits robust performance and increases its performance with utmost level of accuracy.