In-Scene Atmospheric Compensation of MWIR Spectral Imagery

This work describes an atmospheric compensation approach for daytime Mid-Wave Infrared (MWIR) spectral imagery. The approach is demonstrated using synthetic hyperspectral data and synthetic as well as measured multispectral data. The general approach borrows concepts from Long-Wave Infrared (LWIR) atmospheric compensation algorithms that search for emissive pixels to retrieve transmittance and upwelling components of the atmosphere, and applies the temperature-emissivity separation algorithm from FLAASH®-IR. A new approach has been developed to determine the incident illumination onto the scene.