International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part Bl. Istanbul 2004
determined DSM / DEM, reference data from the Bavarian
Survey Administration have been made available by the DLR
Oberpfaffenhofen. The location of the reference areas is shown
in figure 2. For the areas Prien, Peterskirchen, Gars and Taching
laser scanner data with a spacing of 5m and a vertical accuracy
better than 0.5m are available. Each of these areas does have a
size of Skm x Skm. For the 10km x 10km area of Inzell in the
moderate northern part laser scanner data with SZ < 0.5m and
in the mountainous southern part a DEM based on digitised
contour lines from maps 1 : 10 000 with an accuracy of only
SZ=5m and a spacing of 25m has been distributed. The
reference DEM of the 30km x 50km area Vilsbiburg has a
spacing of 50m and a vertical accuracy of 2m.
2. DEM DETERMINATION
The image orientation has been determined with the Hannover
program BLASPO using just the information of the view
direction together with the general orbit information and control
points. The image positions of the control points and some seed
points for the image matching have been measured manually
using the Hannover program DPLX. From the shell of DPLX,
the matching program DPCOR can be started. DPCOR is using
the least squares matching in the image space. The core of this
program was developed by C. Heipke. The matching in the
image space is independent from any orientation information.
Only some seed points (corresponding points in the both
images) are required for the matching based on the region
growing. Also control points can be used as seed points. The
automatic matching has been done for every third pixel with a
window size of 10 pixels x 10 pixels leading to sufficient
independent ground points in a raster of approximately 15m x
30m.
Based on the orientation determined by BLASPO, the ground
points of the DEM points are computed by an intersection. The
orientation by BLASPO will be adjusted in a tangential
coordinate system to avoid the negative influence of the map
projection. In the program COMSPO, the intersection is
followed by a transformation to the map projection, so finally
the height information is available in a chosen national
coordinate system. For a more fast data handling the not totally
regular distributed ground points are interpolated into a raster
arrangement by program LISA.
The so derived height model includes the visible surface of the
objects, that means it is a digital surface model (DSM) and not a
digital elevation model (DEM) of the solid ground. The points
not belonging to the bare ground have to be removed before a
comparison with the reference DEMs. This has been done with
the Hannover program RASCOR (Jacobsen 2001, Passini et al
2002). RASCOR is using a sequence of different methods for
the filtering of a DSM in raster form. The operational use
showed, from a random arrangement a raster arrangement can
be interpolated and this can be analysed by RASCOR with
sufficient results even under the condition of not using the
original data.
RASCOR is analysing the DSM and based on this it is
determining the procedure and tolerance limits automatically
without user interaction. RASCOR starts with an analysis of the
height distribution itself. Based on the structure of the achieved
histogram of height distribution an upper and lower limit of the
accepted height can be identified automatically. This methods
requires flat areas, it does not work in rolling and mountainous
terrain. It is followed by an analysis of the height differences of
440
neighboured points. The accepted height limit of neighboured
points is depending upon the slope and the random errors. With
this method only small objects and the boundary of larger
elements can be eliminated, but it is still very efficient.
Even large buildings can be found by a sudden change of the
elevation in a profile to a higher level and a later corresponding
change down, if no vegetation is located directly beside the
buildings. This method can be used for laser scanning, but it is
not optimal for DEMs determined by automatic image matching
where the buildings are looking more like hills.
Other larger objects not belonging to the bare ground are
identified by a moving local profile analysis; at first shorter and
after this longer profiles are used. The required length of the
moving local profile is identified by an analysis of a sequence
of shorter up to longer profiles. In flat areas the individual
height values are checked against the mean value of the local
moving profile, in rolling areas a linear regression is used, in
mountainous areas polynomials have to be used. It will be
combined with data snooping taken care about a not even point
distribution caused by previously eliminated points. All these
methods are applied in X- and Y-direction. Elements which
have not been removed by this sequence of tests are analysed
by a moving surface which may be plane, inclined or
polynomial. The size of the moving surface is identified by the
program itself by checking the data set with a sequence of cells
with different size. As final test a local prediction can be used,
but it is usually only identifying few points not belonging to the
ground after the described sequence of tests.
In the case of the check for height differences of directly
neighboured points, the upper point will be eliminated if the
tolerance limit will be exceeded. The other methods are using a
weight factor for points located below the reference defined by
the neighboured points. This will keep points located in a ditch
or cutting in the data set. Usually points determined by laser
scanning do not have blunders causing a location below the true
position, but this may happen in the case of a DSM determined
by automatic image matching, justifying a weight factor.
In forest areas at first only the trees are removed by the
program, smaller vegetation is remaining, so a second iteration
is necessary. A second iteration in other cases may remove also
terrain points leading to a more generalised DEM. This may be
useful for the generation of contour lines, but it is not optimal
for the correct description of the terrain.
The derived DEMs have been investigated by a comparison
with the reference DEM. Because of the general different
situation of points located in forest or open areas, a separation
of both terrain types has been made in the used Hannover
program for the DEM analysis DEMANAL. DEMANAL can
use a geo-referenced layer for different terrain types. It also
determines the dependency of the vertical accuracy as a
function of the terrain inclination. It is possible to lefine
tolerance limits for the terrain inclination and the discrey ancies
of the DEM-points. The forest layers have been extracte:' from
the topographie maps 1: 50 000.
3. IMAGE ORIENTATION
The identification of the trigonometric points in the HRS-
images was not so simple. Not in any case the correct location
based on the trigonometric point description could be found.
Inter
The
dista
figut
The
prog
scant
gene
in ad
toget
addit
spee«
with
soutl
avail
4) nc
have
right
area
prob
Figur
The |
follo
SX=(
SX-
are si
this i
10m.
highe
vertic
paral|
direct
The
matcl
point
better
expla
point:
The i
made
