The application of photogrammetry to architecture has been an infrequent
exercise for photogrammetrists whose natural tendency should be to reproduce the
conditions and procedures of aerial photogrammetry in the plotting and drawing of
buildings. Sometimes this is possible, and an entire facade may be photographed in
stereopair with camera axes normal to the facade, with relative orientation following
the procedure for nearly vertical aerial photographs of essentially flat terrain, and
with control measurements on the plane of the facade to determine absolute orientation
and scale. More often than not, however, the building and its location do not permit
this particular photogrammetric procedure. First the camera must be inclined up-
wards. Then in indeterminate seas of grass or blank sky the points are lost at which,
in aerial stereopairs, the elements of relative orientation can be isolated. Finally the
orientation which satisfied the general plane of the facade reveals its weakness when
plotting is pushed in depth to distant towers or to nearby steps and terraces.
It is the need for accuracy in great depth of plotting that places architectural
photogrammetry firmly in the category of terrestrial photogrammetry. The studies
of such men as Professor Bertil Hallert of Sweden of the problems of terrestrial
photogrammetry have considerable application in the photogrammetry of architecture.
Professor Hallert's "Determination of the Accuracy of Terrestrial Stereophotogram-
metric Procedures, "' published in the March,1955 issue of Photogrammetric Engi-
neering, describes the effect of small errors of the convergence of camera axes upon
the plotting of a plane and upon the propagation of error in depth of plotting. I have