Geopositioning precision analysis of multiple image triangulation using LROC NAC lunar images

作者:Liu, B (Liu, Bin)[ 1,2 ] ; Jia, MN (Jia, Mengna)[ 1,3 ] ; Di, KC (Di, Kaichang)[ 1 ] ; Oberst, J (Oberst, Juergen)[ 2,4,5 ] ; Xu, B (Xu, Bin)[ 1 ] ; Wan, WH (Wan, Wenhui)[ 1 ]

PLANETARY AND SPACE SCIENCE

卷: 162  页: 20-30  特刊: SI

DOI: 10.1016/j.pss.2017.07.016

出版年: NOV 1 2018

摘要

With continuing data acquisition from lunar orbiters, many areas of the lunar surface have been observed multiple times under different viewing conditions. This raises an issue regarding how to select the best stereo sets and achieve the highest 3D geopositioning precision in such areas. This paper presents a comprehensive analysis of the geopositioning precision of multiple image triangulation using Lunar Reconnaissance Orbiter Camera (LROC) Narrow Angle Camera (NAC) images. Seven and nine LROC NAC images are used, respectively, for the Apollo-11 and Chang' e-3 landing sites. The photogrammetric methods are developed based on a rigorous camera sensor model and rational polynomial model. Experiments with different combinations of dual images are performed for comparisons at both sites. The results demonstrate that the geopositioning precision, especially the height precision, is improved as the convergence angle increases from near 0 degrees-50 degrees. More importantly, we find that as the convergence angle of a stereo pair increases, the image matching precision decreases approximately linearly, which makes the geopositioning less precise. We also find that the shadow-tip distance and the aspect ratio have roughly linear effect on degrading matching quality. So these two factors also need to be considered when selecting image pairs for stereo mapping. The geopositioning precision is mainly controlled by the convergence angle when it is less than about 10 degrees while the image matching error plays a more critical role when the convergence angle is greater than 10 degrees. Experiments with multiple images indicate that utilizing more images produces higher precision than almost all dual-image models; meanwhile, using fewer images can produce better precision than using all available images together. A progressive selection method is proposed to find the best image combination for maximum precision and effectiveness.

通讯作者地址: Di, KC (通讯作者)

Chinese Acad Sci, Inst Remote Sensing & Digital Earth, State Key Lab Remote Sensing Sci, Beijing 100101, Peoples R China.

地址:

[ 1 ] Chinese Acad Sci, Inst Remote Sensing & Digital Earth, State Key Lab Remote Sensing Sci, Beijing 100101, Peoples R China

[ 2 ] Tech Univ Berlin, Str 17,Juni 135, D-10623 Berlin, Germany

[ 3 ] Univ Chinese Acad Sci, Beijing 100049, Peoples R China

[ 4 ] DLR Inst Planetary Res, Rutherfordstr 2, D-12489 Berlin, Germany

[ 5 ] Moscow State Univ Geodesy & Cartog MIIGAiK, Moscow 105064, Russia