Share Email Print

Optical Engineering

Performance metrics for the evaluation of hyperspectral chemical identification systems
Author(s): Eric Truslow; Steven E. Golowich; Dimitris G. Manolakis; Vinay K. Ingle
Format Member Price Non-Member Price
PDF $20.00 $25.00

Paper Abstract

Remote sensing of chemical vapor plumes is a difficult but important task for many military and civilian applications. Hyperspectral sensors operating in the long-wave infrared regime have well-demonstrated detection capabilities. However, the identification of a plume’s chemical constituents, based on a chemical library, is a multiple hypothesis testing problem which standard detection metrics do not fully describe. We propose using an additional performance metric for identification based on the so-called Dice index. Our approach partitions and weights a confusion matrix to develop both the standard detection metrics and identification metric. Using the proposed metrics, we demonstrate that the intuitive system design of a detector bank followed by an identifier is indeed justified when incorporating performance information beyond the standard detection metrics.

Paper Details

Date Published: 10 February 2016
PDF: 14 pages
Opt. Eng. 55(2) 023106 doi: 10.1117/1.OE.55.2.023106
Published in: Optical Engineering Volume 55, Issue 2
Show Author Affiliations
Eric Truslow, MIT Lincoln Lab. (United States)
Steven E. Golowich, MIT Lincoln Lab. (United States)
Dimitris G. Manolakis, MIT Lincoln Lab. (United States)
Vinay K. Ingle, Northeastern Univ. (United States)

© SPIE. Terms of Use
Back to Top