Autonomous Detection of Particles and Tracks in Optical Images

When optical navigation images acquired by the OSIRIS‐REx (Origins, Spectral Interpretation, Resource Identification, and Security‐Regolith Explorer) mission revealed the periodic ejection of particles from asteroid (101955) Bennu, it became a mission priority to quickly identify and track these objects for both spacecraft safety and scientific purposes. The large number of particles and the mission criticality rendered time‐intensive manual inspection impractical. We present autonomous techniques for particle detection and tracking that were developed in response to the Bennu phenomenon but that have the capacity for general application to particles in motion about a celestial body. In an example OSIRIS‐REx data set, our autonomous techniques identified 93.6% of real particle tracks and nearly doubled the number of tracks detected versus manual inspection alone. Key Points We describe autonomous techniques for the identification and tracking of particles in motion about a celestial body We demonstrate these techniques using images from the OSIRIS‐REx mission to the active asteroid (101955) Bennu In the OSIRIS‐REx dataset, our autonomous algorithms detected 93.6% of real particle tracks, including 244 tracks not identified by manual inspection