Detailed knowledge of the complex forest structures is essential, for example, to improve flood protection with the help of digital models and to investigate the effects of forest fires. Investigations to date have been based on on-site inspections and are thus only possible on a very small scale, making them very time-consuming and cost-intensive.

Our goal is therefore to develop machine-learning methods that can be used to record and document the forest stand and ground-level vegetation as comprehensively and in as much detail as possible. The laser scanning and image data used in this project are acquired using mobile as well as airborne and satellite sensors.

In addition to this project, two other projects at the TU Darmstadt were also selected for the two-year funding; a total of 12 research approaches at Hessian universities are funded Ministry of Science and the Arts.