KIBI aims to simplify the mapping of protected plant communities through machine learning. This will speed up planning processes and improve environmental monitoring, particularly for sustainability purposes. We'll use readily available datasets like multispectral Sentinel-2 images and aerial imagery.

Additionally, we'll incorporate new data generated within the project, such as multispectral images from flyovers.

The aim of the sub-project of the Forschungszentrum Jülich GmbH (IBG-2) is to classify different plant communities from satellite and aerial data using a machine learning method. In addition to the plant community, an uncertainty estimate should also be given, which can be used for further planning. Research questions to be answered during the project include: Which data sources are suitable for the determination of protected plant communities and with which quality they can be determined. Besides the influence of the spatial resolution of different input sensor data, another goal is to identify relevant wavelengths of different sensors.

Contacts IBG-2

Prof. Ribana Roscher