The Geometric Imaging Model for High-Resolution
Optical Remote Sensing Satellites Considering
Light Aberration and Atmospheric
Refraction Errors
Mi Wang, Ying Zhu, Yanli Wang, and Yufeng Cheng
Abstract
With advances in satellite maneuvering imaging capabil-
ity, stereoscopic images with large roll and pitch angles can
be captured to improve the efficiency of observations. At
the same time, the influences of light aberration and atmo-
spheric refraction on image positioning accuracy will be more
significant. However, these errors are not accounted for in the
traditional imaging and calibration model for optical agile
satellites. In this study, the formation mechanisms of the ab-
erration and atmospheric refraction errors in optical remote
sensing satellite Earth observation imaging were analyzed
quantitatively, and correction models were constructed. From
this, the traditional geometric imaging model was refined by
introducing a correction model for aberration and atmospher-
ic refraction errors to create a more comprehensive geomet-
ric imaging model. The feasibility of an extended rational
function model, based on the constructed more comprehen-
sive geometric imaging model, was verified quantitatively.
Introduction
With advances in industrial manufacturing capabilities, in-
creasingly more optical remote sensing satellites are equipped
with agile imaging functionality. At present, the traditional
geometric imaging model and geometric
are based on an ideal collinear equation;
age point, the imaging object point, and t
satisfy the three-point collinearity princi
Wang
et al.
2017; Zhang
et al.
2014). With enhancements in
satellite maneuvering imaging, more and more satellites have
agile imaging functionality. Agile imaging offers both large-
angle and multiangle imaging capabilities. Large-angle imag-
ing can capture images that deviate greatly from the subastral
point (+40–50°). Multiangle imaging captures continuous
images of the same sensitive area. However, the influence of
environmental factors on imaging light propagation increases
with the imaging angle. Compared with traditional three-point
collinear imaging, agile imaging is inevitably affected by at-
mospheric refraction and light aberration (Greslou
et al.
2008;
Greslou
et al.
2012; Noerdlinger 1999). The light aberration
error changes with the imaging angle due to the high-speed
relative motion between the satellite platform and the imag-
ing object point. At the same time, the atmospheric refraction
error also changes with the imaging angle, as the line of sight
(
LOS
) passes through the Earth’s atmosphere. Because the im-
aging angles of traditional push-broom satellites are relatively
stable, the influence of atmospheric refraction and light aber-
ration can be approximated as systematic errors and compen-
sated for by the installation angle; thus, these errors are usu-
ally not taken into account in traditional satellite geometric
imaging models or geometric calibration models (Cao, Yuan,
and Gong 2015; Chen
et al.
2015). However, with satellite agile
imaging, unsystematic errors for multiscene images and the in-
terior of single-scene images caused by atmospheric refraction
and aberration errors increase with geometric resolution and
imaging range. With continuous improvement in satellite at-
titude determination accuracy, the errors of the imaging model
gradually become an important component of the positioning
errors. Therefore, further improvements in image positioning
accuracy of optical remote sensing satellites require that these
errors be compensated for with high accuracy modeling.
In traditional imaging and calibration models, light aberra-
tion and atmospheric refraction errors are neglected or taken
as systematic errors; however, these errors change continu-
ously with the imaging roll and pitch angles (Greslou
et al.
2008; Greslou
et al.
2012; Noerdlinger 1999). The geometric
calibration of high-resolution linear push-broom satellites
usually includes both external and internal calibration sys-
termination of the installation angles of
d the internal distortion of the satellite
he geometric quality of satellite images
y the combination of camera installation
angles and internal distortion parameters (Baltsavias, Li, and
Eisenbeiss 2006; De Lussy
et al.
2012; Delvit
et al.
2012; Di
et
al.
2014; Fraser and Hanley 2003). Therefore, the traditional
calibration model, which doesn’t consider light aberration and
atmospheric refraction errors, can’t also achieve the optimal
calibration effect. Based on the comprehensive analysis of pre-
vious studies, although the light aberration has been discussed
preliminarily in the reference Greslou
et al.
(2012), and the
atmospheric refraction has been discussed preliminarily in the
reference Yan
et al.
(2015), in a more realistic situation, light
aberration and atmospheric refraction errors exist simultane-
ously. Therefore, the more comprehensive geometric imaging
model with light aberration and atmospheric refraction errors
compensation capability is necessary for the optical remote
sensing satellite. This paper is aimed at building this imaging
model and making a more comprehensive analysis.
Mi Wang, Ying Zhu, Yanli Wang, and Yufeng Cheng are
with the State Key Laboratory of Information Engineering in
Surveying, Mapping and Remote Sensing, Wuhan University,
Whuhan 430079, China.
Yufeng Cheng is the correspondence author, and also with
China Aerospace Science & Industry Corp 8511 Research
Institute, Nanjing 210007, China. (
)
Photogrammetric Engineering & Remote Sensing
Vol. 86, No. 6, June 2020, pp. 373–382.
0099-1112/20/373–382
© 2020 American Society for Photogrammetry
and Remote Sensing
doi: 10.14358/PERS.86.6.373
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
June 2020
373
