For those that are interested, hyperspectral analysis is an old research and development field with at least 30+ years of prior art to read through and apply. The biggest issues with getting useful results from hyperspectral--namely characterization and identification of broad spectrum signatures of complex organics--requires careful calibration and ground truth data for a variety of collection scenarios, e.g. weather, time of day, sun angle, look angle, etc. There are entire careers that have been spent optimizing algorithms and procedures to try to tackle these issues. Hyperspectral from space is hard if you want anything other than gas detection (narrow feature signatures) or anomaly detection based on my experience.
For those that are interested, hyperspectral analysis is an old research and development field with at least 30+ years of prior art to read through and apply. The biggest issues with getting useful results from hyperspectral--namely characterization and identification of broad spectrum signatures of complex organics--requires careful calibration and ground truth data for a variety of collection scenarios, e.g. weather, time of day, sun angle, look angle, etc. There are entire careers that have been spent optimizing algorithms and procedures to try to tackle these issues. Hyperspectral from space is hard if you want anything other than gas detection (narrow feature signatures) or anomaly detection based on my experience.