Hyperspectral sensor testbed for real-time algorithm evaluation

Abstract

Hyperspectral imaging has proved to be a valuable tool for performing material based discrimination of targets in highly cluttered backgrounds. A next step for utilizing this technology is to integrate spectral and spatial discrimination algorithms for autonomous target recognition (ATR) applications. This paper describes a hardware and software testbed system for performing spectral/spatial ATR and presents initial results from a field test in the Anza-Borrego desert.