AFSSI-C: The adaptive feature-specific spectral imaging classifier


Journal Article

We have developed an architecture for spectral classification in spectral imaging. Adaptive kernels are designed using probabilistically-weighted principal component analysis. Simulation predicts orders-of-magnitude reduction in error rates, and a prototype system is currently under construction. © 2012 OSA.

Full Text

Duke Authors

Cited Authors

  • Dunlop, MJ; Jansen, PA; Golish, DR; Gehm, ME

Published Date

  • January 1, 2012

Published In

  • Computational Optical Sensing and Imaging, Cosi 2012

Digital Object Identifier (DOI)

  • 10.1364/cosi.2012.cm4b.4

Citation Source

  • Scopus