ibul 2004
ed (Kim,
om some
ind SPOT
ual to 30,
| , 2003).
f SPOT-5
the stereo
d through
iso show
condition
he proper
er various
images in
ve a good
"a SPOT
Journal of
schroeder,
DEM and
data over
ikationen/
vI., 2003.
Different
zes with a
SAR data,
y
Hd, 2001
ScanSAR
Data.
m.
from the
images,
DEM GENERATION FROM HIGH RESOLUTION SATELLITE IMAGERY USING
PARALLEL PROJECTION MODEL
A. Habib, E. M. Kim, M. Morgan, I. Couloigner
Department of Geomatics Engineering, University of Calgary,
Calgary, 2500 University Drive NW, Calgary, AB, T2N 1N4, Canada
(habib@geomatics.ucalgary.ca, emkim@ucalgary.ca, mfmorgan@ucalgary.ca, couloigner@geomatics.ucalgary.ca)
TS: HRS DEM Generation from SPOT-5 HRS Data
KEY WORDS: SPOT, High resolution, Satellite, Push Broom, Modelling, DEM, Stereoscopic, Matching
ABSTRACT:
DEM generation from stereoscopic imagery is contingent on establishing the mathematical model relating the scene coordinates of
conjugate points to the ground coordinates of the corresponding object point. Either rigorous or approximate models can be used to
establish such a relationship. Rigorous modelling necessitates a full understanding of the imaging geometry associated with the
involved sensor. Moreover, it involves the external characteristics (as represented by the Exterior Orientation Parameters — EOP) and
the internal characteristics (as represented by the Interior Orientation Parameters — IOP) of the imaging sensor. Such characteristics
are derived with the help of control information, which might take the form of a calibration test field, ground control points, and/or
onboard navigation units (e.g., GPS/INS). However, the derivation of these parameters might not be always possible due to: the lack
of sufficient control; weak imaging geometry (especially for satellite imaging systems with narrow angular field of view); and/or
intentional concealment by the data provider (e.g., Space Imaging does not release the IOP and the EOP for their commercially
available imagery). Therefore, there has been an increasing interest to investigate approximate models, which do not explicitly
involve the internal and external characteristics of the imaging system. Among them, parallel projection has become a popular model
for its simplicity and accurate representation of imaging sensors with narrow angular field of view, which is the case for newly
launched high resolution satellite imagery (e.g., IKONOS, QUICKBIRD, SPOT-5, ORBVIEW, and EOS-1). This paper presents a
complete methodology for DEM generation from stereo-satellite scenes using parallel projection modelling of the imaging geometry.
The performance of the developed methodology is evaluated through real datasets captured by SPOT-1, SPOT-2, and SPOT-5.
most difficult problems associated with DEM generation from
high resolution imaging satellites. To ensure a reliable solution
DEM generation from remotely sensed imagery 1s crucial for a of the matching problem, prominent features, as represented by
1. INTRODUCTION
variety of mapping applications such as ortho-photo generation, an interesting signal around the selected primitives, are usually
city modelling, object recognition, and creation of perspective used (Tomasi et al., 1991; Harris et al, 1988; Forstner, 1986).
views. Recently launched high resolution imaging satellites Restricting the search space, where conjugate points are sought
(e.g., SPOT-5, IKONOS, QUICKBIRD, ORBVIEW, and EOS- for, is another factor that can be used to reduce matching
1) constitute an excellent source for efficient, economic, and ambiguities. Generating normalized scenes (i.e., resampled
accurate generation of DEM data for extended areas of Earth’s scenes according to epipolar geometry) is the most common
surface. In general, the procedure for DEM generation from approach for restricting the search space for conjugate points
stereoscopic views can be summarized as follows (Shin et al., (Cho et al., 1992). In normalized scenes, conjugate points are
2003): expected to lie along the same row in overlapping scenes.
e Feature selection in one of the scenes of a stereo-pair:
3 ; ; The mathematical relationship between the scene and object
Selected features should correspond to an Interesting
: : coordinates of conjugate points can be established using either
plienomenon m the sons ador tic Ob Set ne rigorous or approximate modelling of the perspective geometry
. Identification of the conjugate feature in the other scene: of the imaging system. Rigorous modelling rognires n
TRIS Brobicm is known as (He matching/corresponde ee comprehensive understating of the imaging geometry. In this
problem within the photogrammetric and computer vision type of modelling, the IOP as well as tie EOP of the imaging
communes se, system are explicitly involved in the mathematical relationship
e Intersection procedure: Matched points in the stereo-scenes between corresponding scene and object coordinates (Habib and
undergo an intersection procedure to produce the ground Beshah, 1998). Deriving such characteristics requires the
coordinates of corresponding object points. Thé ^ availability of control information, which might be in the form
intersection process involves the mathematical model ^ or, calibration test field, ground control points, and/or onboard
relating the scene and ground coordinates. — navigation units (GPS/INS). However, deriving such parameters
« Point densification; High density elevation data is ight be hindered by: the lack of sufficient control; weak
generated within the area under consideration through 41 ^ imaging geometry (e.g. narrow angular field of view of the
interpolation in-between the derived points in the previous imaging sensor); and/or intentional concealment of these
step. parameters by the data provider (e.g., Space Imaging does not
provide the IOP and the EOP for their commercially available
The matching problem and the mathematical model relating the à |
IKONOS imagery). Therefore, there has been an increasing
scene and ground coordinates of corresponding points are the
303
