; a
ric
the
lo-
10-
ng
ge
ati-
nd
try
the
iti-
| photographs or digital images |
point transfer
-»1 mono-
measurement
triangulation 7
/
I
/
1
1
1
4
J
y
exterior
orientations
coordinates of tie points
Figure 4: The process of triangulation and blockadjust-
ment.
4.2 Triangulation and block adjustment
In figure 4 the process of triangulation and block ad-
justment is visualised. For the triangulation step of the
photogrammetric process standards have been for-
mulated for e.g. the number of tie points to be measu-
red and their location in the aerial images in the so-
called Von Gruber locations. The interior and relative
orientation are part of the triangulation. The standards
for the orientation steps are summarized in table 1.
The number of observations has to be at least twice as
large as the number of parameters involved.
The standards for the bundle block adjustment incor-
porate specifications for the functional and stochastic
model to be used. As far as the functional model is
concerned the bundle method is preferred over the
independent model method. The adjustment software
has to include statistical testing in the form of the so-
called data snooping method.
interior | relative | absolute | rectification
# parameters (p) | 4 5 Ts aie iB
# points 4 12 12 | 12
S observations (o) | 8 16 12 36: | 124
redundancy (ofp) | 2 27 24 $453
criterion ( 2 o ) 10 um 10 um 2 V(om2 + 0)
om2 : measuring precision ow : control point precision
Table 1: Criteria for the orientation steps
4.3 Plotting and field completion
The plotting and the field completion are the most
time-consuming steps. Mostly stereo-plotting is used
(depending on the requirements even in the relatively
flat terrain of The Netherlands stereo-plotting is neces-
sary). Stereo-plotting greatly enhances the interpreta-
905
tion and thereby the quality of the mapping. Superim-
position is used to monitor the progress and to check
the completeness of the mapping.
It is advised to register geometric relations such as
parallelity of lines or buildings being square during
plotting. These relations should not be applied in the
plotting phase, but processed in one go with the infor-
mation gathered during field completion. Depending on
the method used for combining the plotting result with
the existing map (the connection of point fields), geo-
metric relations should not be applied as a part of the
field completion, but integrated in the adjustment for
the connection of point fields.
Map revision. For map revision in built-up areas with
flat terrain a monoscopic digital approach can be ad-
vantageous. After scanning the images are rectified
using available natural control points. The criteria for
this procedure are listed in table 1. Mono-plotting is
then combined with super-imposition. This relatively
new approach is very efficient and for certain elements
of the map it meets the accuracy requirements in
many parts of the country.
The pilot project mentioned in section 3 included a test
using mono-plotting in combination with super-imposi-
tion of the cadastral map (Mulder, 1994). Making use
of the existing cadastral map resulted in savings of
some 40%. Of course these savings will depend on
the correspondence between topographic features and
parcel boundaries. The costs roughly equaled the costs
of the same test with stereo-plotting instead of mono-
plotting. The geometric quality of the mono-plotting
was not as good as the quality, of the stereo-plotting;
the GBKN-requirements were not met in this test. The
applied photo scale is one of the reasons (1:5000
instead of 1:3000 in built-up areas and 1:12000 inste-
ad of 1:6000 in rural areas). Height differences could
have played a role as well. It is expected that the
precision can be improved for instance by using a
smaller pixel (after resampling the pixel size used was
30 um).
Field completion. The goals of the field completion
phase of a photogrammetric project are the completion
of the map and a check on the map contents. Field
completion consists of:
- preparation;
- field work;
- updating of plotting results.
For the preparation of the field work (colour) plots are
made. In figure 5 plot samples are shown before and
after field completion. The field work normally includes
the gathering of semantic information and terrestrial
measurements, for example the measurement of roof
eaves. The precision of these measurements should
correspond to the precision of point definition of the
points involved. In this way the precision of the final
map will not degrade (significantly) due to field com-
pletion.
Updating the plotting result includes the processing of
the terrestrial measurements and the integration of the
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B4. Vienna 1996
