ESA GNC Conference Papers Repository

During close approach to a small body, it is necessary to reconstruct the small body shape and to estimate the relative geometry between the small body and the probe. This implies to define the small-body-fixed reference frame, to perform the rotational dynamics characterization, to localize the probe with respect to the small-body-fixed reference frame, and to reconstruct the small body shape. Current approach strongly relies on radiometric tracking and stereophotoclinometry but it is not suited for on-board applications. Thus, autonomous navigation solutions must be developed to enable small body characterization and small-body-fixed navigation throughout the mission. This paper proposes a vision-based solution to perform simultaneous localization and mapping that exploits tracked features and limbs information. On the one hand, the features tracking enables the small-body-fixed frame definition, the rotational dynamics estimation and the smallbody-fixed localization. On the other hand, limbs help constraining the sparse features-based shape reconstruction and computing the final polyhedral shape reconstruction. To test the proposed solution, the Hayabusa close approach to Itokawa is presented and commented to have an overview of the algorithm performances.