The requirement for automated evaluation of detectability of spectral signatures in varying backgrounds is a common theme in many applications of spectral imaging. Existing spectral imaging sensors are quite mature and have been demonstrated in hand held, land based, vehicle mounted, and airborne applications. While these sensors offer high sensitivity resulting in high quality spectra, reliable and consistent, automated detection and identification in realistic scenarios has not been adequately demonstrated. Under this Phase I effort, ASIT will concentrate on developing innovative spectral processing technologies and prototype demonstration of a hyperspectral detectability assessment tool that addresses real-world problems including: time varying spectral mixtures, with unknown background composition, and unknown interferant signatures.
Benefits: To address both Government and commercial needs for enhanced understanding of spectral detection and identification performance, this effort will combine existing processing technology with specialized software