This study examines the issues related to two of the most palmprint promising approaches applied to the contactless biometric authentication and presents a performance evaluation on three different scenarios. The presence of significant scale, rotation, occlusion and translation variations in the contactless palmprint images requires the feature extraction approaches that can accommodate such within class image variations. Therefore the usage and performance of traditional palmprint feature extraction methods on contactless imaging schemes remain questionable and hence all/popular palmprint feature extraction methods may not be effective in contactless frameworks. The experimental results on more than 6000 images from three contactless databases acquired in different environments suggest that the scale invariant feature transform (SIFT) features perform significantly better for the contactless palmprint images than the promising orthogonal line ordinal features (OLOF) approach employed earlier on the more conventional touch-based palmprint imaging. The experimental results further suggest that the combination of robust SIFT matching scores along with those from OLOF can be employed to achieve more reliable performance improvement. The use of publicly available databases ensures repeatability in the experiments. Therefore this study provides a new/challenging contactless hand database acquired in uncontrolled environments for further research efforts.
ASJC Scopus subject areas
- Computer Vision and Pattern Recognition