This is not the latest version of this item. The latest version can be found here.
Deep Learning Crater Detection for Lunar Terrain Relative Navigation
Name
Scitech_Submission_final-1.pdf
Description
Accepted version
Size
3.49 MB
Format
Adobe PDF
Checksum (MD5)
ec012572698d1d177b2dcd6ad02ea995
Date Issued
January 2020
Journal
AIAA Scitech 2020 Forum
Publisher
American Institute of Aeronautics and Astronautics (AIAA)
Citation
2020. "Deep Learning Crater Detection for Lunar Terrain Relative Navigation." AIAA Scitech 2020 Forum, 1 PartF.
Version
Author's final manuscript
Abstract
© 2020, American Institute of Aeronautics and Astronautics Inc, AIAA. All rights reserved. Terrain relative navigation can improve the precision of a spacecraft’s position estimate by providing supplementary measurements to correct for drift in the inertial navigation system. This paper presents a system, LunaNet, that uses a convolutional neural network to detect craters from camera imagery taken by an onboard camera. These detections are matched with known lunar craters, and these matches can be used as landmarks for localization. The motivation for generating such landmarks is to provide relative location measurements to a navigation filter, however the details of such a navigation filter are not explored within this work. Our results show that on average LunaNet detects approximately twice the number of craters in an intensity image as two other intensity image-based crater detectors. One of the challenges of cameras is that they can generate imagery with vastly different appearances depending on image qualities and noise levels. Differences in image qualities and noise levels can occur for reasons such as changes in irradiance of the lunar surface, heating of camera electronic elements, or the inherent fluctuation of discrete photons. These image noise effects are difficult to compensate for, making it important for a crater detection system to be robust to them. Convolutional neural networks have been demonstrated to be robust to these kinds of imagery variation. LunaNet is shown to be robust to four types of image manipulation that result in changes to image qualities and noise levels of the input imagery.
Terms of Use
Creative Commons Attribution-Noncommercial-Share Alike
Persistent DSpace Link
DOI of Published Version
10.2514/6.2020-1838
