@inproceedings{oai:jaxa.repo.nii.ac.jp:00007430,
 author = {吉光, 徹雄 and 久保田, 孝 and So, Edmond Wai Yan and Yoshimitsu, Tetsuo and Kubota, Takashi},
 book = {アストロダイナミクスシンポジウム講演後刷り集, Proceedings of 18th workshop on JAXA Astrodynamics and flight mechanics},
 month = {Mar},
 note = {Symposium on Flight Mechanics and Astrodynamics 2008 (July 28-29, 2008. Institute of Space and Astronautical Science, Japan Aerospace Exploration Agency (JAXA)), Sagamihara, Kanagawa Japan, On the surface of an asteroid, the mobility of a robotic explorer will be affected by the low-gravity environment. Wheeled rovers are expected to be less effective due to reduced traction. Hopping has been proposed as an alternative mechanism for locomotion, because rough terrain can be traversed efficiently., Robotic exploration on the remote surface of an asteroid will require autonomous navigation, for which accurate localization is an essential element. With conventional wheeled rovers, odometry is a simple and widely-used localization technique. However, with a hopping rover, no such inherent method exists. The authors are thus investigating the use of computer vision to perform fast and accurate relative localization., A hopping rover is typically small and compact, with all its instruments enclosed within its body to protect them from impact upon landing. As a result, only a very limited baseline is available for stereo vision. Thus, the authors are focusing on a monocular system. With a single camera, relative motion can be recovered using optical flow, which is part of the Structure-from-Motion (SfM) problem. The slow speed and static environment on an asteroid surface are particularly favorable conditions that should allow this method to produce useful results., 形態: カラー図版あり, Physical characteristics: Original contains color illustrations, 資料番号: AA0064733056},
 pages = {348--353},
 publisher = {宇宙航空研究開発機構宇宙科学研究本部, Institute of Space and Astronautical Science, Japan Aerospace Exploration Agency (JAXA)},
 title = {Relative Localization of a Hopping Rover on an Asteroid Surface Motion Estimation using Optical Flow},
 volume = {2008},
 year = {2009}
}