Share Email Print

Proceedings Paper

3D matching techniques using OCT fingerprint point clouds
Author(s): Henrique S. Gutierrez da Costa; Luciano Silva; Olga R. P. Bellon; Audrey K. Bowden; Raphael K. Czovny
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Optical Coherence Tomography (OCT) makes viable acquisition of 3D fingerprints from both dermis and epidermis skin layers and their interfaces, exposing features that can be explored to improve biometric identification such as the curvatures and distinctive 3D regions. Scanned images from eleven volunteers allowed the construction of the first OCT 3D fingerprint database, to our knowledge, containing epidermal and dermal fingerprints. 3D dermal fingerprints can be used to overcome cases of Failure to Enroll (FTE) due to poor ridge image quality and skin alterations, cases that affect 2D matching performance. We evaluate three matching techniques, including the well-established Iterative Closest Points algorithm (ICP), Surface Interpenetration Measure (SIM) and the well-known KH Curvature Maps, all assessed using a 3D OCT fingerprint database, the first one for this purpose. Two of these techniques are based on registration techniques and one on curvatures. These were evaluated, compared and the fusion of matching scores assessed. We applied a sequence of steps to extract regions of interest named (ROI) minutiae clouds, representing small regions around distinctive minutia, usually located at ridges/valleys endings or bifurcations. The obtained ROI is acquired from the epidermis and dermis-epidermis interface by OCT imaging. A comparative analysis of identification accuracy was explored using different scenarios and the obtained results shows improvements for biometric identification. A comparison against 2D fingerprint matching algorithms is also presented to assess the improvements.

Paper Details

Date Published: 16 February 2017
PDF: 7 pages
Proc. SPIE 10068, Imaging, Manipulation, and Analysis of Biomolecules, Cells, and Tissues XV, 100680O (16 February 2017); doi: 10.1117/12.2253107
Show Author Affiliations
Henrique S. Gutierrez da Costa, Univ. Federal do Paraná (Brazil)
Luciano Silva, Univ. Federal do Paraná (Brazil)
Olga R. P. Bellon, Univ. Federal do Paraná (Brazil)
Audrey K. Bowden, Stanford Univ. (United States)
Raphael K. Czovny, Univ. Federal do Paraná (Brazil)

Published in SPIE Proceedings Vol. 10068:
Imaging, Manipulation, and Analysis of Biomolecules, Cells, and Tissues XV
Daniel L. Farkas; Dan V. Nicolau; Robert C. Leif, Editor(s)

© SPIE. Terms of Use
Back to Top