The paper is about the estimation of the relative position of a spacecraft, during the Entry Descent Landing (EDL) phase, by means of computer vision. A camera installed on board of the vehicle acquires images that are used for estimating the relative position of the camera between two consecutive images. A crucial point of the analysis, and the main objective of this work, is the estimation of the fundamental matrix F, considering the fact that in most cases we deal with a quasi-degenerate configuration. Indeed, the distance between the spacecraft (and the camera) and the planet surface, together with the morphology of the ground, make the problem difficult since most of the points will be extracted from a dominating plane. We discuss two different ways of addressing such degeneracy, while keeping the computational cost low, and present very promising results on synthetic as well as real image sequences. © 2011 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Zini, L., Odone, F., Verri, A., Lanza, P., & Marcer, A. (2011). Relative pose estimation for planetary entry descent landing. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 6469 LNCS, pp. 255–264). https://doi.org/10.1007/978-3-642-22819-3_26
Mendeley helps you to discover research relevant for your work.