the laser
strained
sent, the
e pixels
ed. Line
an point
is easily
methods,
| number
will be
lue, such
'stigated,
f Schmid
segments
nents for
nts.
he stereo
f the line
segments
(s. The
nent with
ched was
hich was
aint, such
terior and
oment are
inates, an
> imagery
rientation.
1993) for
data and
e of these
important
indication
irrounding
; detected,
he ground
> breakline
as, certain
ed that the
lequate for
re, such as
h. the area
ffects of a
Kerry McIntosh
gradual slope, therefore the ground points are below a certain elevation. The higher points are either vegetation points
or building points, or extraneous points such as those on power lines that are not of interest in this research.
The classification approach used to determine
which points are vegetation and which are
building points depends on the spatial
frequency of the data. In high frequency data,
the spatial distribution of the data may be
used in the classification process, however
with lower resolution data it is more difficult
to incorporate this information. "Vegetation
has a random spatial distribution, whereas
building points have a planar distribution, as
shown in Figure 2. As the information being
‘used in this research has a low spatial
resolution, an assumption is made that
vegetation will be isolated points or groups of
isolated points, whereas buildings cover a
larger area.
Vegetation
Building
Figure 2. Laser data showing vegetation, buildings and the
area used for experimentation.
3.2.5 Surface Discontinuities. The most important edges to be detected are the discontinuities between the roof and
the ground. These areas are important to the accurate representation of the visible surface, as they contain dramatic
changes in elevation. Where a roofline has been detected which has no corresponding ground surface breakline, a new
breakline is added to accurately describe the visible surface.
The breakline representing the roof will be projected vertically onto the ground surface to produce a new breakline, thus
describing the surface at ground level as well as roof height. The location of the new breakline must be slightly offset
outward from the roofline, as the triangulation process being used does not allow points to exist with the same
horizontal location, as would be the case for a vertical plane. The elevation of the ground surface at the location of the
added breakline must be determined. It is proposed that the laser data in the areas surrounding the breakline be
searched. The new breakline will be generated and assigned the elevation of the ground points in that area.
3.2.6 TIN Generation. The laser points that occur on the breaklines are eliminated, again due to the triangulation
process not accepting points with identical horizontal locations. The triangulation process is used to generate a TIN that
utilizes the laser data points and constrains the triangles to follow the breaklines. The inclusion of the elevation
information to the triangulation provides the digital surface model produced by combining the laser data and the
photogrammetric data.
4 EXPERIMENTATION AND RESULTS
The initial testing was carried out using a data set over Ocean City, Maryland, USA, which includes digital stereo
imagery and laser data. The data set covers different types of areas, including residential areas, flat terrain, beach front,
dunes, canals and high-rise buildings. Only a small portion of a residential area has been used for the initial testing of
the algorithm. For detailed information regarding the test site, see Csathó er al. (1998).
The laser data have been transformed to the coordinate system of the photogrammetric data using the parameters
determined by the process of surface registration. The two data sets are therefore on the same coordinate system and
represent the same surface. Laser points in the new data set are eliminated if they occur within close proximity to the
breaklines. The breaklines and the laser data are then merged to create a new data set.
4.1 Surface Generation and Registration
The transformation parameters were determined using the developed registration algorithm. Rooflines were measured
analytically to produce planes, and these planes were used with the laser points in the registration process, thus
providing a result that was not dependent on automatic DEM generation techniques. The transformation parameters
between the surfaces were in the order of one to two decimeters. Further details are provided in Schenk er a/. (2000).
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B3. Amsterdam 2000. 567