SAR Imaging and Interferometry Using
Parameters Estimated from Raw Data
Dongliang Wang, Jun Yang, Guangcai Sun, Matthew Brolly, Xin Tao, Jianhua Xiao, Guoqing Sun, Youchuan Wan and Xiaoping Xin
Abstract
Due to the costs, sensitivity, and export policies of many
governments, universities, and research institutions, particu-
larly in developing countries, the ability to purchase, install
and maintain high-accuracy inertial navigation system (
INS
)
and global positioning system (
GPS
) is restricted. This paper
presents a new method for
SAR
imaging and interferometry
using parameters estimated from raw data. First, methods
for determining the order of real and imaginary parts within
the raw data images, and for determining chirp rate polarity
are proposed. Second, a selection of parameters, including
the forward effective velocities of the sensor, the near range
distance, and the squint angle, are extracted using the Doppler
centroid and Doppler rate. Finally, we create single look
complex (
SLC
) images, coherence maps, digital elevation models
(
DEMs
), and dual-pass differential unwrapped phase maps. The
level of accuracy shown in this comparative study suggests that
the proposed method is acceptable for creating the featured
SAR
products and is suitable for real world applications. This
method and result is particularly relevant for systems which
suffer from a lack of high-accuracy positional metadata.
Introduction
Satellite
SAR
platforms generally orbit at heights significantly
above the Earth’s atmosphere and record precise orbit data;
for example, The
ALOS/PALSAR
system orbits at 697 km above
the Earth’s surface. At such altitudes the effects of short-peri-
od motion compensation and orbit errors are less significant
than at lower orbits or airborne flight altitudes. When pro-
cessing satellite
SAR
data in the azimuth direction, the Dop-
pler centroid and the frequency rate may be calculated using
accurate spacecraft ancillary data incorporating information
including the near range distance and velocity. These satellite
GPS
-determined parameters may be directly employed for
DEM
generation (Rossi
et al
., 2012; Martone
et al
., 2012; Crosetto,
2002) and deformation monitoring (Dell’Acqua and Polli, 2011;
Gerke and Kerle, 2011; Jones and Davis, 2011; Kaya
et al
.,
2011; Zhang
et al
., 2012; Gernhardt and Bamler, 2012; Herre-
ra, 2009). For airborne
SAR
platforms, the motion errors can
be considerably higher than those associated with equivalent
spaceborne platforms due to atmospheric turbulence and the
associated aircraft properties such as the relatively small size of
the craft. This makes the need for motion compensation (
MOCO
)
in airborne SAR operations more pressing than required when
using higher altitude SAR systems (Xing
et al
., 2009).
Due to the costs, sensitivity, and export policies, many
universities or institutes in developing countries, including
China, are unable to buy and install the high-accuracy
INS
and GPS systems on airborne
SAR
platforms required for high
accuracy orbit calculations (Cumming and Li, 2007; Kong
et
al
., 2005; Wu
et al.
, 2013; Cao
et al
., 2010). Therefore, more
accurate information regarding the near range distance and
forward velocities must be obtained using alternative methods
to the typical methods employed to estimate values from the
received radar data (Madsen, 1989; Jin, 1986; Bamler, 1991;
Yu and Zhu, 1997; Wong and Cumming, 1996; Bamler and
Runge, 1991; Cumming and Li, 2007; Kong
et al
., 2005; Callo-
way and Donohoe, 1994; Samczynski and Kulpa, 2010; Wahl
et al
., 1994; Berizzi
et al
., 2002; Dall, 1991; Wu
et al
., 2013;
Cao
et al
., 2010). In general it is a difficult undertaking to
ascertain and verify the accuracy of the estimated near range
distance and velocities recorded in airborne
SAR
systems in
the absence of high-precision flight data.
In this study, we analyze the accuracy of the near range
distance, velocities, accelerations, and displacement rates in
the line of sight (
LOS
) direction calculated using data estimat-
ed from a raw data pair of
ALOS/PALSAR
scenes. The data are
accompanied by high-precision orbit data contained within
the recorded meta data files and serve as a direct comparison
to establish the accuracy exhibited by the novel methods
introduced in this work.
Dongliang Wang is with the National Hulunber Grassland Eco-
system Observation and Research Station, Institute of Agricul-
tural Resources and Regional Planning, Chinese Academy of
Agricultural Sciences, Beijing 10081, China and the School of
Remote Sensing and Information Engineering, Wuhan Univer-
sity, Wuhan 430079, China (
).
Jun Yang and Guangcai Sun are with the National Key Lab-
oratory of Radar Signal Processing, Xidian University, Xi’an
710071, China.
Matthew Brolly is with the School of Environment and Tech-
nology, University of Brighton, Brighton BN2 4GJ, UK.
Xin Tao and Guoqing Sun are with the Department of Geograph-
ical Sciences, University of Maryland, College Park, MD 20742.
Jianhua Xiao is with the Wuhan Geomatics Institute, and the
Key Laboratory of Precise Engineering & Industry Surveying
of National Administration of Surveying, Mapping & Geoin-
formaton, Wuhan 430022, China.
Youchuan Wan is with the School of Remote Sensing and Infor-
mation Engineering, Wuhan University, Wuhan 430079, China.
Xiaoping Xin is with the National Hulunber Grassland Eco-sys-
tem Observation and Research Station, Institute of Agricul-tural
Resources and Regional Planning, Chinese Academy of Agri-
cultural Sciences, Beijing 10081, China
Photogrammetric Engineering & Remote Sensing
Vol. 80, No. 7, July 2014, pp. 000–000.
0099-1112/14/8007–000
© 2014 American Society for Photogrammetry
and Remote Sensing
doi: 10.14358/PERS.80.7.000
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
July 2014
663
