We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
2D and 3D palmprint fusion and recognition using PCA plus TPTSR method.
- Authors
Cui, Jinrong
- Abstract
This paper employs both two-dimensional (2D) and three-dimensional (3D) features of palmprint for recognition. While 2D palmprint image contains plenty of texture information, 3D palmprint image contains the depth information of the palm surface. Using two different features, we can achieve higher recognition accuracy than using only one of them. In addition, we can improve the robustness. To recognize palmprints, we use two-phase test sample representation (TPTSR) which is proved to be successful in face recognition. Before TPTSR, we perform principal component analysis to extract global features from the 2D and 3D palmprint images. We make decision based on the fusion of 2D and 3D features matching scores. We perform experiments on the PolyU 2D + 3D palmprint database which contains 8,000 samples and achieve satisfying recognition performance.
- Subjects
PALMPRINT recognition; PRINCIPAL components analysis; IMAGE fusion; BIOMETRIC identification; ROBUST control; HUMAN facial recognition software; ARTIFICIAL neural networks
- Publication
Neural Computing & Applications, 2014, Vol 24, Issue 3/4, p497
- ISSN
0941-0643
- Publication type
Article
- DOI
10.1007/s00521-012-1265-y