International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B5. Istanbul 2004
(DMCTM). The ADS40TM which has been developed in co-
operation with the DLR (Deutsches Zentrum für Luft- und
Raumfahrt — the German Aerospace Centre) is a three-line
pushbroom scanner with 28? forward fore and 14?aft viewing
angles from the nadir. With this design each object point is
imaged 3 times with a stereo angle of up to 42?. Each
panchromatic view direction includes 2 CCD-lines, each with
12000 pixels in staggered arrangement, leading to 24 000
pixels, covering a swath of 3.75km from a flying altitude of
3000m with a 15cm ground pixel size. The DMCTM has a
different design; it integrates 4 digital panchromatic cameras
(and 4 multispectral 3k x 2k cameras) with CCD of 4Kx7K
resolution, resulting in images with 8k by 14k resolution. With a
pixel size of 12um x 12um and a focal length of 120mm, the
camera has a 43.1° by 75.4° field of view. A bundle block
adjustment of a small block with crossing flight directions with
an image scale 1:12 800 (flying height 1500m) resulted in a
sigma0 of 2um (1/6 pixel). At independent check points
standard deviations of o, — o, —/-4cm corresponding to 3.3um
in the image (1/4 pixel) and o, =+/-10cm corresponding to
2.7um x-parallax were achieved (Doerstel et al 2002). Such
accuracy is far beyond what is achievable with a film-based
camera. It should be noted that although digital frame cameras
do currently not reach the accuracy of film-based sensors, other
digital imaging techniques surpass them by a considerable
margin.
Another important development in digital imaging is the
development of airborne hyperspectral imaging sensors. These
sensors are used to map different bands of the visible and
invisible spectrum. Typically they are pushbroom scanners and
can produce more than 100 different bands or channels. The
combination of specific bands produces a unique signature for
each material in the scene. These signatures are used to classify
and identify the materials present at each location. It is
therefore an excellent tool for environmental assessments,
mineral mapping and exploration, vegetation communities and
species, health studies, and general land management studies.
This imagery is especially powerful when combined with
LIDAR points or the LIDAR generated surface. For example
the extraction of forest canopy heights can be accomplished
using a combination of hyperspectral classification and LIDAR
based multiple return analysis techniques.
Among the new technologies, Airborne IFSAR (Interferometric
Synthetic Aperture Radar) mapping is attracting much attention
in the geo-spatial community. This attention is due to the
flexibility of system deployment, near weather-independent
operation, cloud penetrating capability, versatile map products,
and quick turn-around time. As a result, high-resolution
airborne IFSAR systems are providing data to applications
traditionally supported by conventional Photogrammetric
technology. The three main products are, Digital Elevation
Models (DEMs), digital Orthorectified Radar Images (ORRIs),
and Topographic Line Maps (TLMs).
3. DEVELOPMENT OF GEO-REFERENCING
TECHNOLOGY
Direct geo-referencing is the determination of time-variable
position and orientation parameters for a mobile digital imager.
The most common technologies used for this purpose today are
satellite positioning by GPS and inertial navigation using an
Inertial Measuring Unit (IMU). Although each technology can
in principle determine both position and orientation, they are
usually integrated in such a way that the GPS receiver is the
main position sensor, while the IMU is the main orientation
sensor. The orientation accuracy of an IMU is largely
determined by the gyro drift rates, typically described by a bias
(constant drift rate), the short tem bias stability, and the angle
random walk. Typically, four classes of gyros are distinguished
according to their constant drift rate, namely:
l. strategic gyros (0.0005-0.0010 deg/h or degree per month)
2. navigation-grade gyros (0.002-0.01 deg/h or degree per
week)
tactical gyros (1-10 deg/h or degree per hour)
4, low-accuracy gyros (100-10 000 deg/h or degree per
second)
La
Only the last three classes will be discussed in the following.
Operational testing of direct geo-referencing started in the early
nineties, see for instance Cannon and Schwarz (1990) for
airborne applications, and Lapucha et al (1990) for land-vehicle
applications. These early experiments were done by integrating
differential GPS with a navigation-grade IMU (accelerometer
bias: 2-3 10-4ms-2, gyro bias: 0.003 deg/h) and by including
the derived coordinates and attitude (pitch, roll, and azimuth)
into a photogrammetric block adjustment. Although GPS was
not fully operational at that time, results obtained by using GPS
in differential kinematic mode were promising enough to
pursue this development. As GPS became fully operational the
INS/DGPS geo-referencing system was integrated with a
number of different imaging sensors. Among them were the
Casi sensor manufactured by Itres Research Ltd., see Cosandier
et al (1993); the MEIS of the Canada Centre for Remote
Sensing, see ; and a set of CCD cameras, see El-Sheimy and
Schwarz (1993). Thus, by the end of 1993 experimental systems
for mobile mapping existed for both airborne and land vehicles.
A more detailed overview of the state of the art at that time is
given in Schwarz et al (1993). The evolution of the geo-
referencing technology during the past decade was due to the
ongoing refinement and miniaturization of GPS-receiver
hardware and the use of low and medium cost IMU’s that
became available in the mid-nineties. Only the latter
development will be briefly discussed here.
The inertial systems used in INS/GPS integration in the early
nineties were predominantly navigation-grade systems,
typically strapdown systems of the ring-laser type. When
integrated with DGPS, they provided position and attitude
accuracies sufficient for all accuracy classes envisaged at that
time. These systems came, however, with a considerable price
tag (about US $ 130 000 at that time). With the rapidly falling
cost of GPS-receiver technology, the INS became the most
expensive component of the geo-referencing system. Since
navigation-grade accuracy was not required for the bulk of the
low and medium accuracy applications, the emergence of low-
cost IMU in the mid-nineties provided a solution to this
problem. These systems came as an assembly of solid state
inertial sensors with analog read-outs and a post-compensation
accuracy of about 10 deg/h for gyro drifts and about 10-2 ms-2
for accelerometer biases. Prices ranged between US$ 10 000
and 20 000 and the user had to add the A/D portion and the
navigation software. Systems of this kind were obviously not
suited as stand-alone navigation systems because of their rapid
position error accumulation. However, when provided with
high-rate position and velocity updates from differential GPS
(1s pseudo-range solutions), the error growth could be kept in
bounds and the position and attitude results from the integrated
solution were suitable for low and medium accuracy
AA cr n
e
Pn 7
C (Jo
