Share Email Print

Proceedings Paper

Regression techniques for material identification in hyperspectral data
Author(s): Randy S. Roberts
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Identification of materials in hyperspectral imagery is a fundamental analysis task. Materials are often identified by building pixel models using a library of reference spectra along with a regression technique. This paper describes several regression techniques that are useful in modeling hyperspectral pixels, demonstrates the characteristics of the algorithms on simulated data, and compares the strengths and weaknesses of the techniques

Paper Details

Date Published: 18 June 2003
PDF: 8 pages
Proc. SPIE 5001, Optical Engineering at the Lawrence Livermore National Laboratory, (18 June 2003); doi: 10.1117/12.500372
Show Author Affiliations
Randy S. Roberts, Lawrence Livermore National Lab. (United States)

Published in SPIE Proceedings Vol. 5001:
Optical Engineering at the Lawrence Livermore National Laboratory
Theodore T. Saito; Monya A. Lane, Editor(s)

© SPIE. Terms of Use
Back to Top