This project is a collaboration between CVLab, LASTRO, and the EPFL Space Center. A major outcome of 2025 was the completion of a comprehensive database of objects extracted from astronomy pictures, integrating light curve analysis algorithms to support the physical characterization of observed space objects. In parallel, 6D pose estimation algorithms derived from orbital data were developed, advancing the project’s core capabilities for accurate in-orbit tracking and debris capture preparation.
Contacts with Swiss industry partners continued throughout the year, consolidating the transfer of expertise to the Swiss space economy. CVLab pursued its collaboration with software company Klepsydra on the optimization and deployment of computer vision techniques for 6D pose estimation on space-grade hardware. The Space Center further advanced the partnership between Uni Bern, CVLab, and the Swiss space situational awareness (SSA) company s2a-Systems. The object database, built upon satellite and space debris detections extracted from the VST astronomical data archive, now provides a structured repository of detections, identification of known space objects, and lightcurves (time series of brightness measurements) enabling detailed physical characterization. Follow-up projects to expand the data analysis pipeline and incorporate additional astronomical data archives are underway.