Combined Calibration Method Based on Rational
Function Model for the Chinese GF-1
Wide-Field-of-View Imagery
Taoyang Wang, Guo Zhang, Yonghua Jiang, Siyue Wang, Wenchao Huang, and Litao Li
Abstract
The internal geometric distortion of charge-coupled-device
(
CCD
) chips degrades the internal geometric accuracy of satellite
images. Even if a large number of ground control points (
GCPs
)
are used for geometric correction, the satellite images cannot
always be used for direct mapping. Because of the current lack
of core parameters for building a rigorous geometric model,
this paper proposes a combined interior-orientation calibration
method based on the rational function model (
RFM
) for mul-
tiple
GF-1
wide-field-of-view (
WFV
) images of the same general
scene. A polynomial model of look angles in the image space is
chosen to compensate for the systemic error of
CCD
chips. After
calibration for sensors 1 to 4, the orientation accuracy is im-
proved to 0.4 pixels. Moreover, the residual error figures show
no obvious trend. Other
GF-1
WFV
images are calibrated using
the obtained interior-orientation parameters, and the improve-
ment in their orientation accuracy is demonstrated to validate
the effectiveness of the proposed method.
Introduction
The methods for improving the geometric accuracy of satellite
images can be broadly divided into two: interior calibration and
exterior orientation. Methods that rely on interior calibration
are aimed at solving problems in the interior geometry of satel-
lite images caused by lens distortion or the deformation of
CCD
chips. Methods that rely on exterior orientation, on the other
hand, are aimed at solving problems caused by measurement
error in orbit and attitude, as well as camera installation errors,
in the process of satellite imaging. Although laboratory cali-
bration for crucial parameters of satellite cameras is performed
before the satellite launch, deformations in the interior-orien-
tation elements of the satellite can be introduced through the
satellite launch process. A number of other physical factors
(e.g., temperature changes) also affect the satellite imaging sys-
tem, and all these factors can degrade the geometric quality of
the image. Therefore, the distortion of satellite images needs to
be compensated for using in-orbit geometric calibration. After
calibration, improved orientation accuracy can be achieved for
satellite images even with a small number of ground control
points (
GCP
s). Conversely, if there is considerable internal
geometric distortion in the image, adding
GCP
s will not signifi-
cantly improve the orientation accuracy. Thus far, experts and
scholars have performed a significant amount of work regarding
in-orbit geometric calibration of satellite imagery. However, all
these efforts have been based on the rigorous geometric model
(Beckett
et al.
, 2009; Breton
et al.
, 2002; Gachet, 2004; Grodecki
and Dial, 2002; Grodecki and Lutes, 2005; Radhadevi and So-
lanki, 2008; Radhadevi
et al.
, 2011; Tadano
et al.
, 2006; Tadano
et al.
, 2007; Tadano
et al.
, 2009; Zhang
et al.
,2014).
GF-1
is the first satellite of the Chinese high-resolution earth
observation system. Table 1 lists its orbit- and attitude-control
parameters and wide-field-of-view(
WFV
)payload indicators.
T
able
1. GF-1 O
rbit
and
A
ttitude
C
ontrol
P
arameters
and
WFV P
ayload
I
ndicators
Parameters
Indicator
Orbit type
Sun-synchronization repeating orbit
Altitude
645 km (Nominal value)
Orbit inclination
98.0506°
Local time of descending node 10:30 AM
Side swing ability (rolling)
±25°, motor time for 25°
≤
200s,
emergency side swing (rolling) ±35°
Spectral range
0.45 - 0.52μm
0.52 - 0.59μm
0.63 - 0.69μm
0.77 - 0.89μm
Spatial resolution
16 m
Width
800km (combination of 4 cameras)
Revisit cycle (side swing)
4d
Cover cycle (no side swing)
8d
The
WFV
system of
GF-1
is composed of four cameras (see
Figure 1) with a total cross-track width of coverage of up to
800 km (Bai, 2013). Because of the satellite data manage-
ment policy, level 0 satellite image data cannot be obtained.
Therefore, a rigorous geometric imaging model is difficult to
build because of the lack of core parameters (satellite attitude
and orbit parameters, camera installation angle, line integral
time, etc.). In the publicly released
GF-1
level 1
WFV
images,
only the rational polynomial coefficient (
RPC
) file and satel-
lite image file are available for the user. Between level 0 and
level 1 of
GF-1 WFV
images, relative radiometric calibration
is performed, but geometric calibration is rarely performed.
Therefore, the
GF-1
level 1
WFV
images are not regarded to
have geometric calibration. However, owing to the lack of
Taoyang Wang and Yonghua Jiang are with the School of Re-
mote Sensing and Information Engineering, Wuhan Univer-
sity, 129 Luoyu Road, Wuhan, 430079, P.R. China (wangtaoy-
ang@ whu.edu.cn).
Guo Zhang, Wenchao Huang, and Litao Li are with the State
Key Laboratory of Information Engineering in Surveying,
Mapping and Remote Sensing (LIESMARS), Wuhan Universi-
ty, 129 Luoyu Road, Wuhan, 430079, P.R. China.
Siyue Wang is with the School of Electronic Information and
Communication, Huazhong University of Science and Tech-
nology, 1037 Luoyu Road, Wuhan, 430074, P.R. China.
Photogrammetric Engineering & Remote Sensing
Vol. 82, No. 4, April 2016, pp. 291–298.
0099-1112/16/291–298
© 2016 American Society for Photogrammetry
and Remote Sensing
doi: 10.14358/PERS.82.4.291
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
April 2016
291
