Short-wave/Infrared Imagery Fusion Using a Physics-based Model

This proposal addresses the enhanced utilization of space-based Thermal IR (TIR) imagery, a powerful remote sensing method that can operate in nighttime, fog, and haze conditions and can detect subtle thermal signatures. Currently, ambiguities inseparating temperature, emissivity and atmospheric effects complicate the detection and identification of targets against cluttered backgrounds, even using multiple wavelength bands. To address this problem we developed and demonstrated in Phase I thefeasibility of a novel Short-wave/Infrared Fusion Technique (SIFT) for multispectral imagery, in which VIS-SWIR data (which may be from a prior collect and/or on a different platform) are used together with a physics-based temperature and emissivityretrieval algorithm to enhance the information content of TIR images. It enables retrieval of more accurate surface temperatures with minimal contamination from emissivity clutter and retrieval of more accurate spectral emissivities. This will lead toimproved detection and identification of surface materials and objects. In Phase II, a refined SIFT will be coded into a versatile commercial prototype software package, dubbed the MSDA (Multi-Spectral Data Analysis) Toolkit, for data simulation, dataanalysis, and sensor trade studies, and used to analyze a variety of airborne and spaceborne data collects.