Share Email Print

Proceedings Paper

Multispectral imaging for biometrics
Author(s): Robert K. Rowe; Stephen P. Corcoran; Kristin A. Nixon; Robert E. Ostrom
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Automated identification systems based on fingerprint images are subject to two significant types of error: an incorrect decision about the identity of a person due to a poor quality fingerprint image and incorrectly accepting a fingerprint image generated from an artificial sample or altered finger. This paper discusses the use of multispectral sensing as a means to collect additional information about a finger that significantly augments the information collected using a conventional fingerprint imager based on total internal reflectance. In the context of this paper, “multispectral sensing” is used broadly to denote a collection of images taken under different polarization conditions and illumination configurations, as well as using multiple wavelengths. Background information is provided on conventional fingerprint imaging. A multispectral imager for fingerprint imaging is then described and a means to combine the two imaging systems into a single unit is discussed. Results from an early-stage prototype of such a system are shown.

Paper Details

Date Published: 23 March 2005
PDF: 10 pages
Proc. SPIE 5694, Spectral Imaging: Instrumentation, Applications, and Analysis III, (23 March 2005); doi: 10.1117/12.589487
Show Author Affiliations
Robert K. Rowe, Lumidigm, Inc. (United States)
Stephen P. Corcoran, Lumidigm, Inc. (United States)
Kristin A. Nixon, Lumidigm, Inc. (United States)
Robert E. Ostrom, Lumidigm, Inc. (United States)

Published in SPIE Proceedings Vol. 5694:
Spectral Imaging: Instrumentation, Applications, and Analysis III
Gregory H. Bearman; Anita Mahadevan-Jansen; Richard M. Levenson M.D., Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?