ORIENTATION OF SATELLITE AND AIRBORNE IMAGERY FROM MULTI-LINE
PUSHBROOM SENSORS WITH A RIGOROUS SENSOR MODEL
Daniela Poli
Institute of Geodesy and Photogrammetry, ETH Zurich, 8093 Zurich, Switzerland- daniela@geod.baug.ethz.ch
Commission I, WG 1/5
KEY WORDS: Pushbroom, Sensors, Modelling, Orientation, Self, Calibration
ABSTRACT:
Today CCD linear array scanners are widely used on satellite, airborne and helicopter platforms to provide images with along track
stereo viewing. For the orientation of this kind of imagery, models based on the rigorous description of the acquisition geometry, on
rational polynomial functions or on affine transformations are used. An overview is presented. Among these approaches, the model
developed at the Institute of Geodesy and Photogrammetry (IGP), ETH Zurich, belongs to the class of rigorous models and is
applicable to a wide class of pushbroom sensors carried on satellite, airplane and helicopter. The model can be used with single-lens
and multi-lens sensors with synchronous and asynchronous stereo acquisition. The sensor position and attitude are modelled with 2M
order piecewise polynomials depending on time. Additional pseudo-observations allow the reduction of the polynomial order from 2
eX
i
to | if the trajectory allows it. In case of sensors carried on aircraft, the observations from GPS and INS instruments are integrated in
the piecewise polynomials and are corrected from constant shifts ans misalignments between the GPS and INS local systems and the
camera one and systematic errors contained in the observations. A self-calibration is also included for the corrections of radial and
decentering lens distortions, principal point(s) displacement, focal length(s) variation and CCD line(s) rotation in the focal plane.
Using a minimum of 6 Ground Control Points (GCPs) and, additionally, Tie Points (TPs), the external orientation and seclf-
calibration parameters, together with the TPs ground coordinates, are estimated in a least-square adjustment.
In order to demonstrate the model flexibility and potentials, different imagery from pushbroom sensors (TLS, EROS-AI, SPOT-
5/HRS, ASTER, MOMS-02, MISR) have been oriented. In this paper a summary of the results obtained are presented and discussed.
1. INTRODUCTION
CCD linear array sensors, also called linear scanners, are widely
used for the acquisition of images at different ground resolution
for photogrammetric mapping and remote sensing applications.
They scan the ground surface with an array of CCD elements in
pushbroom mode. The image is formed by a side-to-side
scanning movement as the platform travels along its path.
CCD linear array sensors for remote sensing and
photogrammetric applications are usually mounted on aerial and
satellite platforms. Aerial platforms are primarily stable wing
aircraft, but also helicopters are used. Acquisition from airborne
sensors are often used to collect very detailed images and
facilitate the collection of data over virtually any portion of the
Earth's surface at any time. In space, the acquisition of images is
sometimes conducted from the space shuttle or, more
commonly, from satellites for Earth Observation (EO). Because
of their orbits, satellites permit repetitive coverage of the Earth's
surface on a continuing basis. Cost is often a significant factor
in choosing among the various platform options.
The images provided by linear CCD array sensors have very
high potentials for photogrammetric mapping at high and low
scales.
The triangulation and photogrammetric point determination of
pushbroom systems are rather different compared to standard
approaches, which are usually applied for full frame imagery,
(Haala et al., 2000). Additionally, the use of linear imaging
sensors is more difficult with respect to photogrammetric data
processing and requires an increased computational effort
during the subsequent processing chain including matching,
DTM and orthoimage generation.
This work relates to the analysis of the orientation of CCD
linear scanners. Different models can be found in literature. The
rigorous ones are based on the photogrammetry collinearity
equations, which are modified in order to include any external
and internal orientation modelling. Other approaches are based
on rational polynomial functions, affine models and direct
linear transformations. An overview is given in the next section.
At the Institute of Geodesy and Photogrammetry (IGP), ETH
Zurich, a rigorous sensor model for the georeferencing of
imagery acquired by multi-line CCD array sensors, carried on
airborne or satellite, has been implemented. The model fulfils
the requirement of being as flexible as possible and being
adaptable to a wide class of linear array sensors. In fact
pushbroom scanners show different geometric characteristics
(optical systems, number of CCD lines, scanning mode and
stereoscopy) and for each data set specific information are
available (ephemeris, GPS/INS observations, calibration, other
internal parameters). Therefore the model needs to be dependent
on a certain number of parameters that change for each sensor.
The results obtained with different satellite sensors will be
presented.
2. OVERVIEW OF EXISTING MODELS
For the georeferencing of imagery acquired by pushbroom
sensors many different geometric models of varying complexity,
rigor and accuracy have been developed, as described in
(Fritsch et al., 2000, Hattori et al., 2000 and Dowman et al.,
2003). The main approaches include rigorous models, rational
Internati
polynom
affine prc
The rigo
the sensc
which ar:
pushbroc
specific s
used for
spacebor
In case c
followed
called
photogra
sensor e
derived 1
variation
quadratic
added: tl
point co
paramete
formulati
model).
supplenx
is availat
unknown
orientatic
(Baltsavi
Landsat
advantag
pushbroc
paramete
extended
The prin
geometri
imagery
method i
al, 1992
the so-c:
images tl
using La
orientatic
All unki
adjustme
determin
points wl
In the g
model of
spacebor
on MOM
1995), H
model is
imagery,
spacebor
paramete
points, w
100th re
paramete
Lagrange
orientatic
e.g. acq
equation:
attitude «
block-im
depender
additiona
each ext
