Skip to main content

Training Material for automatically deriving training areas for updating land cover classifications

Point of contact
Ursula Gessner
German Aerospace Center (DLR)
Münchener Straße 20
82234 Weßling
Phone: +49-8153-281250


Under the FPCUP action " Downstream service / application development for national statistics and reporting" the German Aerospace Center (DLR) developed a tailor application to the requirements of the German Federal Agency of Cartography and Geodesy (BKG) in collaboration with the German Federal Statistical Office (Destatis) for supporting updates land cover/use classifications using the example of ecosystem classification.

The focus of this tool lies on improving the generation and selection of training data for classification purposes by exploiting Copernicus Sentinel data, in combination with existing geodata.

On the 23rd of November a Python / Jupyter Notebook tool for automatically selecting training areas for supervised land use / land cover classifications was presented at a workshop organised by DLR. Within the tool, the selection of training areas can be based on existing geodata and Sentinel data.

You can find the tool, the presentations and all of the other materials by clicking here.