Comparison of crater-detection algorithms for terrain-relative navigation

Svenja Woicke, Andres S. Moreno Gonzalez, Isabelle El-Hajj, Jelle W.F. Mes, Martin Henkel, R. S.D. Autar, Robert A. Klavers

Research output: Chapter in Book/Conference proceedings/Edited volumeConference contributionScientificpeer-review

8 Citations (Scopus)

Abstract

Precise landings on other bodies require more than just dead reckoning using an inertial measurement unit on-board the lander. If navigation of the lander with respect to a planetary surface is desired, so-called crater detection and crater-matching algorithms might be a valuable asset to find the inertial position of the vehicle using terrain relative navigation techniques. This would enable landing close to an inertially defined landing site, which could, for example, be a surface asset of a previous mission. With the desire to reduce the landing ellipse size, more precise knowledge of the inertial state of the lander is required. Based on an extensive literature review, six different algorithms were implemented to assess the performance of these. This assessment will aid the selection of crater-detection techniques for future precision landing missions. To compare the different algorithms trade-off criteria have been established. The following criteria are assessed: 1) True detection rates 2) False detection rates 3) Accuracy: as the reference maps usually have rather high resolution, inaccuracies of just a few pixels can cause large errors. 4) Run-time: the algorithm should be on-board capable. Moreover, the robustness of the algorithms was investigated. It was found that all algorithms are capable of performing the task of extracting sufficient craters for localising the landing vehicle with respect to a surface map. A method based on extracting and clustering lit pixels delivered the most promising results for the overall detections, whereas, the machine-learning based algorithms showed slightly better robustness.

Original languageEnglish
Title of host publicationAIAA Guidance, Navigation, and Control
PublisherAmerican Institute of Aeronautics and Astronautics Inc. (AIAA)
Number of pages12
Edition210039
ISBN (Electronic)9781624105265
DOIs
Publication statusPublished - 1 Jan 2018
EventAIAA Guidance, Navigation, and Control Conference, 2018 - Kissimmee, United States
Duration: 8 Jan 201812 Jan 2018
https://doi.org/10.2514/MGNC18

Conference

ConferenceAIAA Guidance, Navigation, and Control Conference, 2018
CountryUnited States
CityKissimmee
Period8/01/1812/01/18
Internet address

Fingerprint Dive into the research topics of 'Comparison of crater-detection algorithms for terrain-relative navigation'. Together they form a unique fingerprint.

Cite this