ter
4
iter
> >
o
For a first check on the observations the mean square error of the mean is calculated
from a least 3 independent observations for each point. The r. m.s. ofa single observation was
$ 0. 003 to * 0.004 mm ; points with a r. m.s. larger than * 0. 01 mm were eliminated and printed
out for re-observation.
After adjustment of the relative orientation the v, values calculated with (4) were checked
and for this purpose the vector Vy. was retained in the memory. Points with Vyi? 9 99 (o9 "T.I. s.
of unit weight) were eliminated and printed out for further investigation. The elimination of points
at this step necessitated of course the re-computation of the adjustment. Depending on the position
of the point eliminated in the list of data, the adjustment only or the entire relative orientation is
repeated. Provisions had to be programmed also for the case that a control point has been elimi-
nated. The list of data is reorganized whenever a point has been dropped.
After the step of absolute orientation the residual errors of the control points were checked
according to criteria drawn up for this purpose. Due to differences in quality of the control points,
different criteria were used for a transformation into the photogrammetric control point system
(see also [1] ) and a transformation into the coordinate system of the building site.
The observations from oblique photographs were treated in the same manner and trans-
formed into the photogrammetric control point system utilizing the data of outer orientation -
tilt and swing - recorded at the time of photography.
Mistakes in numbering were located by calculating the length of all meshes of the net and
by automatic plotting of plans and profiles. Experience showed that the graphical depiction of the
results of the measurement is a very sensitive indicator of errors.
Conclusion
The analytical close-range photogrammetric method employed for the evaluation of the
olympic roofs had to be developed under unfavourable circumstances. The short period of time
allotted to the photogrammetric work made it impossible to fully investigate all phases of the pro-
cess. Nevertheless the results of the photogrammetric evaluation could be used for the purposes
of transferring the models of the roofs into reality. Investigations into aspects of analytical close-
range photogrammetric methods are part of a government sponsored research programm in Cons-
tructive Engineering, presently concerned with extended light weight structures, and the photo-
grammetric evaluation of models is one of the main topics of the program.
Literature
[1] Béttinger, W.-U. - "On some Aspects of Taking Close-Range Photographs for Photogram-
metric Evaluation ; Practical Experiences in Photographing the Models of the Cable-Net
Roofs for the Olympiad at Munich". Invited Paper, Com. V, ISP, Ottawa 1972.
[2] Harley, I. A. - "The Calibration of Cameras for Nontopographical Photogrammetry".
Presented Paper, Symposium Com. V ISP, Tokyo 1966.
[3] Schut, G. H. - "Analytical Aerial Triangulation'". Report of the National Research Council
of Canada, Ottawa 1957.
[4] Faig, W. - "Vermessung dünner Seifenlamellen mit Hilfe der Nahbereichsphotogrammetrie''.
Deutsche Geodátische Kommission Bayerische Akademie der Wissenschaften, Reihe C,
Heft 144, Beck'sche Verlagsbuchhandlung München 1969.
[5] Linkwitz, K., Preuss, H.D. - "Die photogrammetrische Vermessung der Modelle der
olympischen D&cher München". Bildmessung und Luftbildwesen Heft 4, 1971.
[6] Linkwitz, K., Preuss, H.D. - Sonderforschungsbereich 64 "Konstruktiver Ingenieurbau'
Annual Report 1970, not published.
[ 7] Sonderforschungsbereich 64 - Annual Report 1971. University of Stuttgart, in print.
[8] Argyris, J.H. et al. - "Messung raumlicher Verformungen mittels fotogrammetrischer
Methoden und elektronischer Datenauswertung'. Institut für Statik und Dynamik der Luft-
und Raumfahrtkonstruktionen, ISD-Report No. 118, Stuttgart 1972.
—D8-
