15 and 
trated 
ervals 
and y- 
  
a dis- 
of the 
. Con- 
f each 
nding 
jen be 
| 
0.22 | 
0.25 | 
0.04 
llaxes 
es, its 
INTRODUCTION 
The programme of work of the Sub-commission IV:4 can be summarized as follows. 
Tests of the fundamental operations of photogrammetry in connection with the controlled experi- 
ments of Sub-commissions IV:1—IV:3, see [1]. Practical tests of the expressions for error propagation 
from the fundamental operations to various results of the photogrammetric procedure, see [2]. 
The fundamental operations can be defined, in accordance with [1], as follows. 
1. The photography 
2. The reconstruction of the ideal bundles of rays from photography 
3. The relative orientation 
4. The absolute orientation and the coordinate determination. 
Owing to the practical conditions of the controlled tests, broadly speaking, only operations 2 and 3 
could be treated by Sub-commission IV:4.! 
In a preliminary paper: General points of view on investigations of the accuracy of photogrammetric 
procedures, the principles of the tests to be performed were described. The paper was distributed to 
all participants in the controlled experiments of Sub-commissions IV:1—IV:3. 
For the test of fundamental operation No. 2, grid measurements were suggested according to a 
well-defined procedure (see below). Such measurements were recommended before and after the 
controlled experiments. 
For the test of fundamental operation No. 3, y-parallax measurements in completely oriented models 
were recommended according to a well defined procedure and in certain model points (see below). 
From the results of the measurements of grid coordinates and y-parallaxes some types of systematic 
errors in the fundamental operations can be determined and the standard error of unit weight of the 
basic measurements can be estimated according to the method of least squares. The accuracy to be 
expected of the final results of the photogrammetrie procedure, primarily the coordinates and eleva- 
tions on the ground, can also be estimated and expressed as root mean square values? of the 
discrepancies to be expected in planimetry and elevation. The estimations are to be checked by 
comparison with the corresponding mean square values of the true discrepancies as determined by 
Sub-commissions IV:1—IV:3. 
The theoretically derived expressions for the error propagation require in principle that only 
accidental (normally distributed) errors shall be present in the fundamental operations and that the 
discrepancies in conditions prevailing, e.g. the relative orientation and redundant control points, 
shall be correctly adjusted. Obviously these requirements will not as a rule be strictly satisfied in the 
1 Rather comprehensive tests of the accuracy of aerial photographs have been performed under different condi- 
tions. Some results will be reported to Comm. I, ISP, for the congress 1960 in the paper: Results of Practical In- 
vestigations into the Accuracy of Aerial and Terrestrial Photographs. See further below under point 2.4. Detailed 
tests of the separate components of the fundamental operations should be performed in the indicated manner. It 
has for instance been proved that the position of a diapositive during the drying process may have a considerable 
influence upon the basic geometrical quality of the image. 
2 Below, the simplified expression mean square value will be used.