On September 22, 2026, Henning Riecken will present his internship results on ” Internship at the German Remote Sensing Data Center (DFD) of the German Aerospace Center (DLR)” at 12:00 at the seminar room 3 in John-Skilton-Str. 4a.
From the abstract: Through the Bavarian Fruit Tree Pact, the Bavarian government, in collaboration with civil society groups, has set the goal of preserving existing fruit tree stands in Bavaria and planting 1 million new fruit trees by 2035. But how can current and future stands be reliably quantified to monitor the implementation of this goal? One possible approach could be the automated detection of fruit trees from Bavarian aerial images using a machine learning workflow. As part of my internship at DLR, I created an XG-Boost and a Random Forest workflow for this purpose. The results showed that, where sufficient training data is available, the models can detect fruit trees with sufficient reliability (F1 score: 0.77). However, the models could not be successfully applied to other areas without such training data (F1 score: 0.06).
1st supervisor: Dr. Sarah Schönbrodt-Stitt Host: DLR
From the abstract: Through the Bavarian Fruit Tree Pact, the Bavarian government, in collaboration with civil society groups, has set the goal of preserving existing fruit tree stands in Bavaria and planting 1 million new fruit trees by 2035. But how can current and future stands be reliably quantified to monitor the implementation of this goal? One possible approach could be the automated detection of fruit trees from Bavarian aerial images using a machine learning workflow. As part of my internship at DLR, I created an XG-Boost and a Random Forest workflow for this purpose. The results showed that, where sufficient training data is available, the models can detect fruit trees with sufficient reliability (F1 score: 0.77). However, the models could not be successfully applied to other areas without such training data (F1 score: 0.06).
1st supervisor: Dr. Sarah Schönbrodt-Stitt Host: DLR
