Cloud cover is the most challenging problem encountered in optical satellite imagery. It leads to difficulty in information extraction, as well as demand for larger constellations, higher revisit times, and larger data storage capacity.
The ultimate goal of CloudlessEO is to propose a way to increase the usability of acquisitions from optical satellites already in orbit by addressing the cloud-cover problem. CloudlessEO is interactive, on-demand, web-based software that consists of a novel generative adversarial network architecture. It exploits the synergies of SAR and optical data from Sentinel-1 and -2 to generate highly accurate synthetic optical data to fill data gaps related to to clouds and their shadows. This will allow the reliable and continuous monitoring of changes in Earth’s surface regardless of location, weather or season.
Furthermore, CloudlessEO gives the user full control over the data sources and methods that are used to form cloud-free images of their areas of interest, along with metrics and suggestions to guide their decision processes.


“We chose CloudlessEO because it addressed our challenge fully and it also has a working implementation that is already serving customers. It has the potential to enable us to do things that were not possible before due to lack of data. Synthetic satellite imagery opens up many possibilities for using both open and commercial data. It is also a perfect fit with our platform and marketplace approach. The solution can be packaged and made available on our marketplace so that all customers can benefit from it.”
António Almeida, Senior Tech Evangelist, UP42


Thetaspace GmbH
Zayd Mahmoud Hamdi