M I DA A EE A CE RNIN 
duced to the arithmetic mean (of all photo-over-. 
laps) of the systematic errors in the image plane. 
For that average of two-dimensional errors in a 
two-dimensional distribution a graphical and an 
algebraic method is developed, the meaning of 
which exceeds the application in photogram- 
metric block adjustment. With the aid of both 
methods, partly novel flying dispositions for 
compensation of systematic image errors are 
conceived; under economic aspects, too, manifold 
flying represents a real alternative to photo-flying 
practised till now. 
The first paper of the session was given by Dr. 
H. Ebner from the Federal Republic of Germany. 
Invited Paper 
Author: Dr. H. Ebner (F.R.G.) 
Title: Self-Calibrating Block Adjustment 
IP 111104 
Dr. Ebner discussed a method for simultaneous 
self-calibration for bundle or independent models 
block adjustment and results obtained with the 
Oberschwaben test site. Results agree with 
theoretical developments based on random errors 
only and confirm the effectiveness of compensa- 
tion of the systematic errors by self-calibration. 
Also in agreement with theory, the bundle ad- 
justment with self-calibration had proved superior 
to independent models with self-calibration. 
In the course of the general discussion that 
followed, Mr. G. Schut (Canada) said it would be 
interesting to make a comparison between the 
correction parameters suggested by the author, 
Mr. de Masson d’Autume and himself. He also 
warned against the use of too many parameters, 
as according to his experience 12 parameters 
double the computing time required. Using ad- 
ditional parameters in only one adjustment saves 
computing time. 
Dr. Ebner answered that the use of several para- 
meters and groups of parameters is justified if 
high accuracy is required. If error detection is 
taken into consideration, the computing time 
needed is increased by only 20 to 30 per cent. 
Presented Papers 
Authors: Dr. P. Wiser (Belgium) 
Dr. F. Ackermann (F.R.G.) 
Title: The OEEPE Test "Oberschwaben" 
PP 111/62 
Dr. Ackermann presented results using the bundle 
adjustment with added parameters in a triangu- 
lation of the Oberschwaben test site. With self- 
calibration, the bundle adjustment furnished the 
more accurate results when compared with an 
independent model adjustment. He observed that, 
in this test, the accuracy obtained was three times 
the one expected, and that in fact the theoreti- 
cally expected accuracy has been reached for 
the first time here. 
Author: Dr. G. Kupfer (F.R.G.) 
Title: On Accuracy Achieved by Different 
Triangulation Procedures PP 111/33 
By order of the State Survey Office of Northrine- 
Westphalia the Rheinbach area near Bonn of 
approximately 8x10 square kilometers had been 
flown twice in perpendicular directions, using a 
at a ne rec centres permet ou ne acc 
a 
N 
LEERE MD A 
Zeiss RMK A 15/23. Before and after that the 
Rheidt test field area had been flown with four 
test strips each for the purpose of partial field 
calibration of the system. 
Dr. Kupfer discussed accuracies achieved by 
different triangulation methods. He found the 
bundle and anblock methods using image data 
refined by field calibration provided results su- 
perior to those obtained with unrefined data. 
The largest corrections of the image coordinates 
were 10 um. 
Author: Dr. W. Tegeler (F.R.G.) 
Title: Densification of Trigonometric Net- 
works by Bundle Adjustment PP /Il/ 
58 
Dr. Tegeler had used a bundle adjustment with 
added parameters in tests made for photogram- 
metric network densification using the Hordorf 
test area. Substantial reductions in the errors 
had been achieved by utilizing the added para- 
meters. 
In 1974 the Landesvermessung of Lower Saxony 
had carried out a photogrammetric network den- 
sification on a photo scale of 1:6000 in the test 
area (6 kmx12 km) of Hordorf. The planimetric 
accuracy of the control and check points had 
amounted to 1 cm. The RMS errors of coordinates 
of the multiple blocks with 10 horizontal control 
points at the perimeter had been as follows: 
Double-blocks: men = + 3,3 cm (5,4 um) 
Fourfold-block: men = + 2,5 cm (4,2 um) 
Author: Mr. K. Jacobsen (F.R.G.) 
Title: Test Kôningshügel, Control by 
Bundle Adjustment PP 111/28 
Mr. Jacobsen had used a bundle adjustment 
with added parameters in Test Kónigshügel. The 
results had been computed with an independent 
model adjustment corrected by prediction with 
filtering. The test area control had had distortions. 
By manipulating the weights of control points, 
the bundle adjustment had produced improved 
results which had been comparable to results 
obtained by prediction and filtering. 
The result of the measurement with the Wild A 7 
had shown a 47 % to 79 % higher error than the 
results of the measurements with the comparator. 
Wednesday, July 14, 1976, from 9.00 to 10.30 
Session 3 of Commission III 
Topic: Aerial Triangulation (3: Methods, 
Computer Programs, Application 
Chairman: Dr. F. Ackermann (F.R.G.) 
Dr. Ackermann opened the meeting and then 
briefly discussed a number of Presented Papers 
that he considered of interest, but for the full 
presentation of which there would not have been 
enough time. 
Author: Colonel G. Agarwal (India) 
Title: A Study of Errors in Slotted Templet 
Combination PP 111/101 
The planimetric accuracy expected of a slotted 
templet combination is usually calculated by an 
empirical formula known as Trorey's formula. 
The applicability of this formula has been ana- 
lyzed with the help of a large statistical sample 
of combinations carried out under controlled 
99