Towards image processing of reentry event

Abstract The space debris population is constantly increasing, including large spare upper stages and spacecraft. Atmospheric re-entry events caused by these artificial objects pose a threat to the ground-based population and their understanding and accurate modelling are required. On the 24th of October 2020, two cameras of the AMOS system in the Hawaiian Islands recorded the re-entry event of spare rocket’s upper stage. In the presented work, we aim to improve and automate the process of fragment identification between simultaneously acquired video recordings of this event. We are presenting the whole processing pipeline. First segmentation is performed to identify the cluster of interest and flying fragments within that cluster in all frames. The acquired information is used to create tracks of individual flying fragments – frame objects to be further associated between recordings.