Biometrics based on multispectral skin texture

  • Authors:
  • Robert K. Rowe

  • Affiliations:
  • Lumidigm, Inc, Suite, Albuquerque

  • Venue:
  • ICB'07 Proceedings of the 2007 international conference on Advances in Biometrics
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

Multispectral imaging (MSI) of the skin provides information about both the surface and subsurface characteristics of the skin tissue. These multispectral characteristics may be described as textures and used to determine the identity of an individual. In this investigation, a multi-day, multiperson study was conducted to compare the performance of a matcher based on multispectral texture analysis with a conventional minutiae-based fingerprint matcher. Both matchers used MSI data from exactly the same area of the sensor. Performance of the two methods was compared for a range of simulated sensor areas. The performance of the textural matcher was nearly equivalent to that of the minutiae matcher for the larger sensor areas within the limited scope of this study. For small sensor areas, the performance of the texture matcher was significantly better than the minutia matcher operating over the identical small region.