system 
ftware 
tially an 
a model 
d to the 
ilable in 
function 
building 
le ina5 
provide 
the site 
ing the 
When placing the order it was specified at which particular 
height the buildings should be registered and with what 
degree of accuracy. As a result, a building that is higher 
than 3m and has a ground area of more than 50 m? should 
be registered as a single object. Parts of buildings that fulfil 
these requirements should also be treated as single 
objects. 
In how far these details have been recognized can only be 
established with the vectors. These will be placed over the 
ortho-photos and can then be checked and compared 
visually. 
The use of the photogrammetric software ORTHOMAX 
makes it possible to show stereopairs 3-dimensionally and 
to measure individual heights with a measuring scale. In 
this way individual building height can be determined. 
However, due to the data available, it is only possible to 
determine the building heights on a random basis. 
4.3 Analytical vs digital photogrammetry 
The composition of the digital building datasests is 
achieved by using two methods. On one hand conventional 
analytical methods were used, on the other the data was 
compiled with the help of the digital method. The digital 
stereo photogrammetric method differs compared to the 
analytical method because a number of processing steps 
run automatically. The most significant part hereby is the 
automatic correlation for determining 3-dimensional rasters 
with various degrees of density. Inspite of this, manual 
adjustments are still neccesssary, especially for the 
determination of the building outlines. That means that with 
both methods the quality of the building outlines 
(completeness, detail and area accuracy) largely depends 
on the individual engineer in charge. Examinations of the 
datasets show that the quality of the data does not vary 
significantly. 
With the analytical method the determination of individual 
buildings heights is done by an engineer who establishes 
the building height by stereopair. It is the decision of the 
engineer at which particular points of the building's roof the 
measuring scale is set. Using the digital method, depending 
on the actual method used (middle value, maximum value), 
the height is determined by the area of the building from the 
raster (Guretzki, Erhardt 1996). 
Due to the actual data available (approx. 60,000 single 
objects) the control of the building heights can only be 
established on a random basis. An engineer establishes 
individual building heights on the digital workstation and 
compares them with the height rasters. This leads to 
problems particularly when complex buildings are involved, 
because in the end it is also the decision of the engineer in 
charge at which point the measuring scale is set. An 
analysis of 100 height points using both methods shows 
that there is a variation up to 5 m in the set points 
established by the engineers. The standard deviation using 
the analytical methods is about 2 m, compared to about 3 
m using the digital method. The accuracy level achieved is 
sufficient for the applications mentioned above. 
5. UPDATING PROCESS 
To use the building data in the microcell planning the actual 
status of the data is significant. Particularly in cities like 
185 
Berlin where acute building activities are carried out, it is 
important to update the available datasets within very short 
periods of time. To make these updates possible a number 
of methods are available. 
5.1 Regular aerial survey 
The most accurate method of updating the data is regular 
aerial surveys. However, as the microcell planning is 
steadily on the increase, it will become necessary to obtain 
building data for other cities so that a complete analysis of 
already existing city data would considerably exceed the 
permitted cost budget. A complete internal evaluation of 
the data within the office is not possible due to a lack of 
resources. House internal surveys can only be limited to 
smaller areas whereby the question arises as to which 
areas should actually be updated. 
In some branch offices of Mannesmann Mobilfunk regular 
evaluation journeys are made to determine the 
field-strength available. All engineers involved possess 
exact knowledge of the situation in the individual cities. 
This information should be used to evaluate part areas. If 
contemporary aerial survey data is available for a particular 
area this should be obtained, otherwise aerial survey flights 
have to be initiated. For these small areas involved a 
subsequent in-house-evaluation can be carried out with the 
existing photogrammetric workstation. 
5.2 Use of other data 
À further possibility is the use of already available digital 
data. Most local construction offices are already in the 
process of using digital data for administration purposes. 
By enquiring regularly at the construction offices areas with 
a high level of building activities can be identified and 
updated if required. 
6. SUMMARY 
At present controls of 3D-buildings can only be carried out 
manually. The possiblity to use photogrammetric 
worstations provides a great help in accomplishing this 
task. However, there still remains the need to have a more 
automated process. The same applies for the updating 
process. Here as well things can only be achieved by a 
higher engagement of engineers and resources. 
References: 
Cichon, D.J.,Kürner, Th.,Wiesbeck, W., Modellierung der 
Wellenausbreitung in uitbanem Gelánde, In: Frequenz, vol. 
47, no.1-2,pp. 2-11, January 1993 
Feistel, M., Baier, A., 1995, Performance ofa 
three-dimensional propagation model in urban 
environments, Sixth IEEE International Symposium on 
Personal, Indoor and Mobile Radio Communications 1995, 
Vol. 2, pp 402-407 
Guretzki, M., Erhardt, H., 1996, Erfassung städtischer 
Gebáudehóhenmodelle mit Einsatz der digitalen 
Photogrammetrie, Bildverarbeitung und ARC/INFO, ESRI 
User Conference Germany 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B2. Vienna 1996