Funding Body: German Federal Ministry for Economic Affairs and Energy (BMWi)
Duration: 2019 to 2022
Budget: 249.000 Euro
Role: Co-PI and main author of proposal
The aim of the AI4Sentinels project is the development of innovative approaches for the processing of Earth observation data from the Copernicus satellite missions Sentinel-1 and Sentinel-2 for large-scale mapping and monitoring applications. In particular, the potential of combining data fusion and domain-specific deep learning approaches will be explored. Specifically, the project will develop software solutions for the conversion of Sentinel-1 SAR images into artificial optical images and the computational removal of clouds from multi-spectral Sentinel-2 images. The resulting novel data products will help end users to implement tasks of continuous monitoring or step-by-step change detection in a more flexible and robust manner, mitigating potential without data loss.
- Meraner A, Ebel P, Zhu XX, Schmitt M (2020) Cloud removal in Sentinel-2 imagery using a deep residual neural network and SAR-optical data fusion. ISPRS Journal of Photogrammetry and Remote Sensing 166: 333-346
- Ebel P, Meraner A, Schmitt M, Zhu XX (2021) Multi-sensor data fusion for cloud removal in global and all-season Sentinel-2 imagery. IEEE Transactions on Geoscience and Remote Sensing: in press