🔎 Our projects #2: Collecting field data 🌳 for #remotesensing 🛰️ based models is time consuming & expensive. In the @DLR funded project UAVforSAT, we @KIT_Karlsruhe & our partners LUP in Potsdam use #deeplearning to extract reference data from very high-resolution #UAV-imagery.
With this reference data, we can then train satellite-based 🛰️ models at landscape-level. For example the fractional cover of standing #deadwood at the Saxon Switzerland National Park.
