A big congratulations to Janik Hoffmann, who successfully defended his Master thesis today on the cutting-edge topic:
“Deep Learning-driven SAR Image Super-Resolution.”
Supervised by Prof. Dr. Tobias Ullmann and Dr. Wolfgang Koppe (Airbus Defence and Space GmbH), Janik’s thesis sits at the intersection of Earth observation and artificial intelligence. His work explores the exciting potential of Convolutional Neural Networks (CNNs) with residual learning to enhance the spatial resolution of TerraSAR-X StripMap imagery, using High-Resolution SpotLight data as a reference.
In an innovative twist, Janik applied deep learning architectures originally developed for natural RGB images to Synthetic Aperture Radar (SAR) data—a notoriously complex but immensely valuable source for environmental monitoring, crisis response, and even military surveillance.
🔍 Key Highlights of His Thesis:
-
Demonstrated that super-resolution techniques can double the spatial resolution of SAR StripMap images.
-
Achieved clear improvements in PSLR (Peak Sidelobe Ratio) and other quality metrics.
-
Verified that super-resolved images significantly increase the number of detectable homologous points in stereo SAR datasets.
-
Opened new doors for Ground Control Point (GCP) retrieval, a key task in SAR-based mapping.
His findings show that deep learning is not only applicable to SAR—but could revolutionize it. Bridging the gap between StripMap and HR SpotLight imagery could greatly enhance future missions by enabling both broad spatial coverage and high detail, simultaneously.
🎯 As SAR super-resolution gains traction in Earth observation research, Janik’s work marks a meaningful step toward integrating deep learning into operational SAR workflows—especially for high-precision applications like GCP detection.
🌐 Janik, we’re incredibly proud of your achievement and the innovative spirit you bring to the field. Your work embodies what the EAGLE MSc is all about: pushing boundaries, challenging norms, and advancing the future of geoscience.
Well done, Janik—and we wish you all the best in your next steps, wherever your journey in science and innovation takes you!
