Our signature project in remote sensing was a project to do terrain classification to avoid helicopter brownout - basically trying to pick landing zones that won't kill the pilots.
This all started with Col Mitch 'Rico' Rios, who brought the problem to NPS in 2003. He started a long line of USMC helicopter pilots who analyzed different ways of approaching the problem of understanding landing zones. After a fair amount of effort, the problem was reduced to a problem in characterizing good and bad landing zones. (including work with commercial systems by Anthony W. Davis, 2007; and Christine Rabaja in 2009). This concluded in some sense with a MERIT program that involved NPS and NRL (Ralph Fiedler). The algorithms we developed were transitioned to the MI 513th, and have proven valuable for operations.
Polarimetric radar image (VV polarization) after new Brownout Algorithm threshold values are applied. Red = Brownout likely; Sienna = Unknown/Ambiguous; Green = Non-Brownout. From Rabaja (2009)