The working group Technology Transfer for Remote Sensing deals with the transfer of methods and tools developed in the section up to an operational usage. For this purpose, the competences of scientific method development are combined with sustainable software development and aspects of business development. The Helmholtz Innovation Lab FERN.Lab is part of the work group.
- Application-oriented methods and software development in remote sensing (FERN.Lab)
- Support of the Section's scientists in transfer-oriented cooperation projects (applications, exploitation plan)
- Support of the Section's scientists in the field of sustainable software development (repositories, licensing)
- Contact point for external cooperation requests from industry, public administration and NGOs
FERN.Lab is the GFZ innovation and technology platform for application-oriented, transdisciplinary method developments for the analysis of remote sensing data. FERN.Lab provides direct access to scientific expertise for the development of remote sensing methods at the Helmholtz GeoForschungsZentrum Potsdam. The experts at FERN.Lab and the ralted scientist of the department have many years of experience in the development of methods and analysis of satellite-based earth observation data. This includes high-resolution optical and thermal remote sensing, radar remote sensing, satellite-based earth gravity field surveying and also high-precision positioning using global satellite navigation systems (GNSS).
In the CropClass project, an operational web service will be developed by combining optical data and radar data. This will allow a highly accurate classification of the main agricultural crops typical of Germany, namely wheat, rye, barley, rapeseed, potatoes, maize and sugar beet, at various times during the growth period of agricultural crops. The classification, based on the combination of the different data sets, is performed using machine learning algorithms.
The basis of the system is the FERN.Lab portal, which enables automated data pre-processing and synergetic data use. Based on this data, AI procedures are evaluated and trained so that a fruit species classification with up to 90% accuracy is aimed for.
Cooperation partner: Dida Datenschmiede GmbH