Simulated recovery of LEO objects using sCMOS blind stacking

We present the methodology and results of a simulation to determine the recoverability of LEO objects using a blind stacking tech-nique. The method utilises sCMOS and GPU technology to inject and recover LEO objects in real observed data. We explore the target recovery fraction and pipeline run-time as a function of three optimisation parameters; number of frames per data-set, exposure time, and binning factor. Results are presented as a function of magnitude and velocity. We find that target recovery using blind stacking is significantly more complete, and can reach fainter magnitudes, than using individual frames alone. We present results showing that, depending on the combination of optimisation parameters, recovery fraction is up to 90% of detectable targets for magnitudes up to 13.5, and then falls off steadily up to a magnitude limit around 14.5. Run-time is shown to be a few multiples of the observing time for the best combinations of optimisation parameters, approaching real-time processing. & COPY; 2023 COSPAR. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).