without having the requirements to cut the original negatives
on the film roll. The scanning can take place in unattended
mode. The modular design of SCAI allows its connection to
various workstations. Scanning does not require special
photogrammetric knowledge and so can be done by non-
photogrammetrists.
The digital and automatic aerial triangulation is the youngest
member in the family of DP-applications (Mayr 1995, Braun et
al. 1996). It is basically possible to fully automatically tie
together blocks of images of any constellation and number of
images. This triangulation task is covered by PHODIS AT.
Sub-dividing in more convenient sub-block sizes is available.
Amongst other topics, special attention was paid in PHODIS
AT to interactive block preparation, the overall user interface,
and the correlation-based ground control point measurement
capabilities.
The automatic generation of DTMs has been in daily use for
several years. In order to improve this process, a special user
interface was added. PHODIS TS (Dorstel 1995) expands the
DTM matching capabilities with stereoscopic superimposition
of the derived DTM and interactive DTM editing.
Digital orthoprojection helped introduce DP sucessfully.
Digital orthophotos are imported into GIS and as such form the
basis of a hybrid data capture method which is called on-
screen digitizing. In addition to digital orthophotos, called
orthoimages, PHODIS OP also delivers radiometrically
balanced mosaics of orthoimages.
The functionalities of a digital stereoplotter as a 3D-digitizer
are explained in more detail in the next chapter. Special
emphasis is placed on those properties which are, from the
point of view of the authors, of particular relevance to GIS.
More literature on the PHODIS ST digital stereoplotter is
found in (Dórstel and Willkomm 1994, Dórstel 1995).
Monoplotting or on-screen digitizing has been practiced in
photogrammetry and GIS for some time. It allows data
acquisition by means of digitizing on top of an orthoimage
which possesses a simple relation to the world cordinate
system, e.g. UTM. The orthoimage is then said to be geo-
referenced. Geo-referenced topographic data sets can be
superimposed on orthoimages and be reviewed e.g. for
changes, in an interactive manner. GIS with on-screen
digitizing capabilities is called hybrid GIS since, as well as
general purpose information and vector data, the GIS now also
deals with the raster data set of the orthoimage. If a DTM is
available and in geo-reference to the orthoimage then a hybrid
GIS can generate and store on-line an elevation for every
planimetrically determined location as a function of its easting
and northing coordinates.
In addition to all of above mentioned functionalities DP offers
the application of automated measurement procedures.
Examples of such are the fully automatic restoration of the
interior orientation of an aerial image (AIO), see (Schickler
and Poth 1996), or the fully automatic measurement of
conjugate image points in order to derive the parameters of the
relative orientation (ARO), see (Tang et al. 1996). Such
modules are essential intrinsic components of the
photogrammetric basis of DP and thus part of PHODIS Base
which acts as a server to the applications on top of it. Open
interfaces for data exchange and programming allow users and
developers to communicate with different software packages.
This property is used in the combination of DP with GIS.
3 GIS and Digital Photogrammetry
Topographic data acquisition is gaining in importance for GIS,
It is well understood that data acquisition is the most cost
effective part of running GIS. Therefore it can be accepted for
economic reasons to integrate photogrammetric data capture
principles when it comes to the modelling of topography in
GIS. This is valid if the usage of such a data acquisition system
is reliable and efficient for the GIS-user. Another important
issue for GIS is that by using DP, available remotely-sensed
data can be processed geometrically. DP takes care of the
specific geometry models underlying various remote sensing
satellites which may produce images in the visible and / or
invisible wave spectrum.
This paper deals with aspects of topographic data acquisition
in and for GIS. It especially concentrates on 3D-data
acquisition. So far topographic 3D-data were compiled
primarily on analytical stereo plotters. These were equipped
with data capture software especially tailored towards the
requirements of the photogrammetric community. Data were
transferred via translator programs from photogrammetric data
bases to the GIS data base. This process caused expensive time
consumption and sometimes also loss of information. In order
to perform 3D-data acquisition easier and more efficiently it
appears natural to apply GIS itself for the topographic data
acquisition task. Missing photogrammetric functionalities e.g.
rectangular closing of houses, compilation of parallel lines by
either digitzizing or by automatic line following, or the
automatic extraction of houses will be thrown in by DP as
selectable modules.
In the 2D-domain GIS already is used for interactive on-screen
digitizing in e.g. ATKIS (Dasing 1995, Kophstahl 1995).
Automating this area is still in the research field (e.g. Heipke
et al. 1994, Englisch 1995, Mayer 1995). Approaches have
been made in semi-automatic and automatic data acquisition.
An example of this is the line following process in large scale
orthophotos. The preferred procedure is the semi-automatic
one (Englisch 1995, Mayer 1995) which appeares to deliver
more reliable results in practical tests. It is interesting that
these algorithms work on underlying orthoimages. These offer
the geo-reference in an easy way. Semi-automatic procedures
require interaction with the user who, for example, sets the
starting points of a variety of processes for extracting houses or
line following. These processes are then invoked. They in tum
will guide the user to all houses and lines found for the
purposes of validity checking and / or entry of additional
information needed to continue. Such a procedure has at least
two advantages. Number one advantage is the fact that the user
does not have to digitize all objects of interest himself. The
DP-system takes over the measurement part of this task. Only
the identification of the objects to be measured has to done by
the user. Number two advantage is that the measurements are
objective measurements.
The expansion of existing GIS to the third dimension has been
done mostly by attaching a DTM to the GIS. In order to fully
integrate the third dimension, some people (e.g. Pilouk and
Kutouiyi 1994, Kraus 1995) propose the use of vertical data
bases. These appear to guarantee the topology of the data and
guarantee unambiguity of data also in the vertical direction. In
general unambiguity is a constraint for 3D-data in GIS.
When the GIS is used directly as the data acquisition system,
data transfers from the photogrammetric data base to the data
base of GIS disappear. This eliminates in an elegant way the
554
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B4. Vienna 1996
attra
supe
pres:
3.1 (
Initi:
mod
COITE
posit
float
the ,
float
imag
resul
base:
It is
the r
calle
corre
impl
three
corre
visua
can 1
guar:
accey
that |
P-mo
OneE
The :
User
corre
and 1
