At the GIL Conference in Stuttgart, University of Hohenheim latest research and developments from science and industry were presented and discussed with special focus on the headline topics of biodiversity and digital agriculture.

Vista presented the lecture “Learning from hyperspectral remote sensing data for machine learning algorithms in earth sciences – A case study to detect plastic-covered greenhouses and plastic-mulched farmlands”. In this presentation, an innovative approach was shown for the generation of essential but often scarce training data from hyperspectral satellite imaging developed in the HyperPin project.

On site our colleague Lena Brüggemann had the chance to discuss about new opportunities for sustainable agriculture and biodiversity, especially enabled through image analysis and AI, and to learn about critical aspects and challenges of the knowledge transfer in digital farming. In terms of AI, different methodologies for automated image analysis and object recognition were shown.