3.3. 
4.1. 
5. 
2:1. 
marked limits of property, 4. purely numerical cadaster, the processing of photogrammetric data 
being done with automatic computers. 
The successful attempt of Fagerholm and Thunberg, to survey from low altitude in spite of poor 
visibility (clouds), and still get accurate results, opens up new, important and wider possibilities 
to photogrammetry. 
Photogrammetric mapping of cities (A/C/IV/2) 
Urban photogrammetric mapping applies mainly where limited but urgent needs occur. Whenever 
time and money are available and a well-established municipal surveying organization, duly inte- 
grated with the juridical and engineering system of the city, exists, the conventional methods can 
be more rational in the long run (at least if the resulting documents are constantly updated). The 
horizontal standard error of well identified plotted points varies beween +5 cm and 4-10 em for 
photo scales of 1:5000 through 1:10 000. In order to make the city map complete and true to 
nature, it is necessary to make either a total previous identification on the photographs, including 
direct positioning of invisible objects, or complementary field work after the plotting, in order to 
fill in the blanks. It is not yet easy to decide, for a given type of city and degree of land occupation, 
which one of these two methods is more suitable. To set up standards in this respect would be 
one of the main goals of research in urban photogrammetry. There is an appreciable influence of 
the plotting machine operator’s personality (skill, conscientiousness, measuring acuity) on the 
quality of the resulting maps. 
Photogrammetric mapping in small scales (A/C/IV/3) 
The percentage of the earth’s surface for which cartographic surveys are needed is still so large, 
that one has to look for procedures simpler, faster, more economical than the ones involving 
terrestrial or aerial triangulation (Panel 4). This applies especially where scales 1:100 000 through 
1:200 000 are desired. Such procedures use for instance airborne means of orientation, greater 
flying heights, super wide-angle cameras, automatic plotting (Stereomat), and third order plotters 
operated by moderately expensive people trained as apprentices. The discussion should show the 
most promising means. 
Inquiries and supervised tests show the following facts: 
— that the horizontal accuracy obtained with photogrammetry is always superior to the 
drafting accuracy, 
- that in order to obtain the desired vertical accuracy one has to use to the limit the possi- 
bilities of photogrammetry, in spite of the small scales involved, 
> that the effort devoted to a developed photo-interpretation or identification determines in 
a decisive manner whether the resulting map can be satisfactory or not as a complete repre- 
sentation of the terrain. 
There are important developments going on with respect to automatic correction for earth curva- 
ture and atmospheric refraction, as well as reduction of the personal vertical errors of operators. 
The use of anaglyphs proves less accurate and slower than the direct stereoscopic observation. 
The possible gain of accuracy secured with modern blink methods (PG Kern) has still to be ascer- 
tained, as well as the advantage offered by such revolutionary innovations as automatic plotting 
of altitudes, contours and profiles. The mistakes and imperfections found in maps are often due 
to injudicious application of methods and instruments, rather than to the instruments or materials 
themselves. 
  
  
  
  
6.1. 
1.1. 
8.1. 
9.1.