A finger-based identification system

This paper presents a multimodal biometric identification system based on the features of human fingers. We describe a new biometric approach to personal identification using the Regularized-Direct Linear Discriminant Analysis (RD-LDA) features extracted from four digitprints (prints of fingers between the first and third phalanx) and four fingerprints, with fusion applied at the matching-score level. The identification experiments were conducted on a database consisting of 1840 hand images (184 people). A correct identification rate of 100% suggests that the system can be used in medium/high-security environments.