Proposed scheme for palm vein recognition based on Linear Discrimination Analysis and nearest neighbour classifier

Palm vein recognition is a new promising field in biometrics. The palm vein pattern provides highly discriminating features that are difficult to forge because it resides underneath the palmar skin. However, the issues of extracting the palm vein features and the high dimension of the feature space are still open. Therefore, in this paper, we propose an improved scheme of palm vein recognition method based on the Linear Discrimination Analysis (LDA) to extract the discriminative features with low dimension. LDA is later followed by the matching procedure using cosine distance nearest neighbor classifier. The performance of the proposed scheme produced 99.50% for identification rate, 100% for verification rate and 0.0% of Equal Error Rate (EER). The experiments prove that the proposed method has a better performance compared with Principal Component Analysis and Gabor filter methods.