High-Resolution Large-Area Digital Orthophoto
Map Generation Using LROC NAC Images
Kaichang Di, Mengna Jia, Xin Xin, Jia Wang, Bin Liu, Jian Li, Jianfeng Xie,
Zhaoqin Liu, Man Peng, Zongyu Yue, Jia Liu, Ruilin Chen, and Changlu Zhang
Abstract
The Chang’e-5 mission of China is planned to be launched
in 2019 to the landing area near Mons Rümker located in
Oceanus Procellarum. Aiming to generate a high-resolution
and high-quality digital orthophoto map (
DOM
) of the planned
landing area for supporting the mission and various scien-
tific analyses, this study developed a systematic and effective
method for large-area seamless
DOM
production. The map-
ping results of the Chang’e-5 landing area using over 700
Lunar Reconnaissance Orbiter Camera (
LROC
) Narrow Angle
Camera (
NAC
) images are presented. The resultant seamless
DOM
has a resolution of 1.5 m, covers a large area of 20° in
longitude and 4° in latitude, and is tied to
SLDEM2015
. The
results demonstrate that the proposed method can reduce the
geometric inconsistencies among the
LROC NAC
images to the
subpixel level and the positional errors with respect to the
reference digital elevation model to about one grid cell size.
Introduction
Mapping of the lunar surface using orbital imagery is one of
the fundamental tasks of almost every lunar orbiter mission.
Among the common mapping products, digital orthophoto
maps (
DOMs
) are essential for measuring and characterizing
lunar surface features. Thus, they are usually used as the base
map for morphological and geological analysis (Wu
et al.
2014; Zhang
et al.
2016; Yue
et al.
2017). H
high-precision
DOMs
are particularly impor
lander and rover missions in terms of landi
safe landing, and surface operations.
China started the Lunar Exploration Program in 2004,
which consists of orbital, soft lander/rover, and sample return
missions (Ouyang
et al.
2004). The first two phases were
achieved by the Chang’e-1, Chang’e-2, and Chang’e-3 mis-
sions, and the third phase (i.e., sample return mission) will
be realized by the Chang’e-5 mission in 2019. The Chang’e-5
mission aims to return about 2 kg of lunar soil and rock
samples. Its target area is near Mons Rümker located in Ocea-
nus Procellarum, which is a large area of lunar mare on the
northwest region of the Moon (Gbtimes 2017). High-resolu-
tion and high-precision mapping of the landing area is critical
to support overall mission planning and detailed analysis of
potential sampling sites (Haase
et al.
2012; Wu
et al.
2014;
Kokhanov
et al.
2017).
So far, there are a number of lunar global orbital image
mosaic maps available, such as Clementine global mosaic
(100 m/pixel) (Robinson
et al.
1999), Lunar Reconnaissance
Orbiter Camera (
LROC
) Wide Angle Camera (
WAC
) globe mosaic
(100 m/pixel) (Arizona State University [ASU] 2011; National
Aeronautics and Space Administration [
NASA
] 2011; Wagner
et al.
2015), Chang’e-1
CCD
camera global mosaic (120 m/pix-
el) (Li
et al.
2010), and Chang’e-2
CCD
camera global mosaic (7
m/pixel) (Data Publishing and Information Service System of
China Lunar Exploration Project 2018; Li
et al.
2015). How-
ever, these products are not sufficient for detailed landing site
analysis due to their low resolutions.
The highest-resolution lunar orbital imagery is achieved by
the
LROC
Narrow Angle Camera (
NAC
), and the images covered
nearly the entire lunar surface at a resolution of 0.5–2.0 m.
However, there are only a limited number of high-resolution
featured mosaics (Klem
et al.
2014) and digital elevation
model (
DEM
) products (Tran
et al.
2010; Burns
et al.
2012) that
have been released by the
LROC
team. The number of images
used in most featured mosaics is from two to tens of
NAC
images. The largest
LROC NAC
mosaic at present is the
LROC
Northern Polar Mosaic (
-
pan;
NASA
/Goddard Space Flight Center/ASU 2014; Wagner
et
al.
2015, 2016), which contains 10 581
NAC
images, covering
°
N to the north pole at a resolution of 2 m.
ent geometric inconsistences of up to ~7 pix-
ed due to problems of the designed procedure
2015). Nevertheless, this product does not
cover the Chang’e-5 planned landing area. The spatial cover-
age of publicly available high-resolution
DOMs
is very limited.
Overall, it is highly desirable to develop effective techniques
for large-area high-precision seamless
DOM
generation.
Photogrammetric processing of high-resolution orbital
images for lunar surface mapping has been performed us-
ing different software systems or methods, such as the US
Geological Survey (
USGS
) Integrated System for Imagers and
Spectrometers (
ISIS
),
SOCET SET
,
NASA
Ames Stereo Pipeline
(Moratto
et al.
2010), and in-house–developed methods. Tran
et al.
(2010) and Burns
et al.
(2012) generated digital terrain
models (
DTMs
) by
LROC NAC
stereo images with the
SOCET SET
software. The resultant
DTMs
have a typical spatial sampling
of 2 m and a vertical precision of 1–2 m (Burns
et al.
2012);
the root mean square (RMS) residuals are typically ~0.25
pixels for a pair of
NAC
stereo images and less than 1 pixel
for multiple sets of stereo images (Tran
et al.
2010). Klem
et
al.
(2014) produced a controlled mosaic for
LROC NAC
images
using the
USGS
ISIS
software, and the seam precision for the
mosaics was generally within 7–8 pixels. Lee
et al.
(2012) and
Kaichang Di, Bin Liu, Zhaoqin Liu, Man Peng, Zongyu Yue,
Ruilin Chen, and Changlu Zhang are with the State Key
Laboratory of Remote Sensing Science, Institute of Remote
Sensing and Digital Earth, Chinese Academy of Sciences,
Beijing 100101, China (
).
Mengna Jia, Xin Xin, and Jia Liu are with the State Key
Laboratory of Remote Sensing Science, Institute of Remote
Sensing and Digital Earth, Chinese Academy of Sciences,
Beijing 100101, China; and the University of Chinese
Academy of Sciences, Beijing 100049, China.
Jia Wang, Jian Li, and Jianfeng Xie are with the Beijing
Aerospace Control Center, Beijing, 100094, China.
Photogrammetric Engineering & Remote Sensing
Vol. 85, No. 7, July 2019, pp. 481–491.
0099-1112/19/481–491
© 2019 American Society for Photogrammetry
and Remote Sensing
doi: 10.14358/PERS.85.7.481
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
July 2019
481
