ime with
cy;
otionless
set level
S of its
maximal
distance
velocity
cle own
VLA
alysis a
]. 590.
Doppler
p pulse
bul 2004
SIMULATOR OF INTERFEROGRAM FOR SPACEBORNE SAR SYSTEM
ab x mn b, + + :
K. Ren? *, G, Wu^ *, X.Q.Shi*, and V. Prinet"
* Dept. of Electronic Engineering, Nanjing University of Sciences and Technology, Nanjing, China-kren@nlpr.ia.ac.cn
E . = T ^ Dee n. . ~ . x ~ .
> National Laboratory of Pattern Recognition, Institute of Automation, CAS -gwu@nlpr.ia.ac.cn
KEY WORDS: SAR, Interferometer, DEM, Remote Sensing, Satellite, Geometry
ABSTRACT:
In interferometric synthetic aperture radar (InSAR) processing, simulation of interferogram is a common practice. It is used as
synthetic data to test and validate the whole chain of INSAR processing from the interferogram creation to the DEM reconstruction.
The objective of this paper is to develop a simulator for validation of geocoding processing and phase-to-height conversion
processing algorithm.
The simulator includes two parts: digital elevation model (DEM) of terrian simulation and interferogram simulation with DEM and
satellite orbit parameters. DEM is realized by the fractal Browian motion (fBm) model with the midpoint displacement method and
the terrain roughness and average slope are determined by two describing parameters of the model. Simulators for interferometry are
generally developed in a simplified imagery geometry model. Here the geolocation method is adopted and the geographic
coordinates (latitude, longitude) of the synthetic DEM in the imagery region are assigned according to the real orbit. Simulation
experiments on simulated and ERS-1/2 tandem real orbit data demonstrate the efficiency of the proposed simulator. The developed
simulator can also be used to test the accuracy of basline estimation.
1. INTRODUCTION
Since increasing spaceborne synthetic aperture radar (SAR)
images over the most part of world is nowadays widely
available, SAR Interferometry (InSAR) technology, as a new
application of SAR, has been an important observation
measurement of Earth in remote sensing community in recent
ten years.
In application research of SAR or InSAR, simulators are
developed for selecting an optimum imagery mode of SAR,
understanding the effects of illumination angle and terrain relief
on SAR images, testing and optimizing interferometric SAR
(IBSAR) processing algorithms, or going insight into radar
received signal of given the terrain [Wary L. S., etc, 2000].
These simulators can be generally divided into two groups. One
is to simulate SAR raw data based on the backscatter model of
land surface and SAR pulse transformation function
[Franceschetti G., 1998]; and the other one is to simulate
interferogram based on a digital elevation model (DEM)[Xu
W.and Cumming B., 1997]. Here we pay attention on the
interferogram simulator. With a high quality DEM as the
simulator input, the topographie phase contribution can be
isolated completely in a real interferogram from other phase
variables caused by object deformation or temporal
decorrelation. When input DEM is coarse, the synthetic
interferogram is useful for flat phase removal processing and
improving the phase unwrapping processing.
In this paper a simulator of interferogram for spaceborne SAR
system, combined with terrain simulation, is developed to
validate phase-to-height conversion and geocoding processing
algorithm. DEM is realized by the fractal Browian motion (fBm)
model with the midpoint displacement method and the terrain
* Corresponding authors.
roughness and average slope are determined by two describing
parameters of the model. For the interferogram simulation, an
improved method is developed, which works with two real
radar sensor parameters and their warp relationship of
coregistration. The simulator implementation is described in
detail. Experiment results on real senor and orbit parameters are
given to demonstrate the efficiency of the presented method.
This paper is organized as follows: Sections 2 and 3 describe
DEM simulation and interferogram simulation respectively.
Section 4 shows the simulation results and Section 5 is
conclusion.
2. DEM SIMULATION
2.1 Principle
In this paper, the terrain data are simulated by fractal Browian
motion (fBm). A fractal is the shape made of parts similar to the
whole in some ways; it can describe many complex objects too
irregular to be dealt with in traditional geometrical language. It
has been shown that terrain surface possesses some fractal
characteristics under a wide range of scale, and fBm is regarded
as a proper model for terrain representation [Peitgen Heinz-Otto
etc., 1988], [Jin Y.W. and Lu S.J., 1998]. FBm is a random
process, whose increment is stationary, and satisfies Gaussian
distribution as follows:
N(0,07) (1)
lax]
