International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B5. Istanbul 2004 
  
  
    
    
   
    
  
  
    
  
  
   
  
  
   
  
  
  
  
  
  
  
   
  
  
   
  
  
  
   
   
  
   
  
  
   
   
   
  
  
  
  
  
  
   
   
  
e The classical photogrammetric method 
(photogrammetric camera Photheo 19/1318 was 
used), 
e The new photogrammetric method (with the digital 
non-metric Kodak DCS 760 camera). 
All acquisitions were made from the same distance of about 45 
m. In the geodetic survey and the classical photogrammetry 
instance, instruments (theodolite, camera) were set on a tripod 
at known coordinates and at a known camera orientation 
element. In the case of the new photogrametric method, the 
picture was taken manually with unknown camera orientation 
elements. 
Figure 7: Picture taken on a test field with the Kodak DCS 760 
digital camera. 
The scale factor for the Photheo 19/1318 camera was m, = 230 
and the scale factor for the Kodak DCS 760 camera was m, 
900. This means that the precision of the Photheo camera is 900 
: 230 = 4 times better. 
There were two versions of the classical photogrammetric 
method. In the first version, a traditional measurement of glass 
plates on a stereo-comparator was carried eut. In the second 
version, scanning of glass plates had been done before the 
measurement of images on the digital station. 
In table 1, accuracy characteristics of pictures are put together. 
According to table 1, we can see that the classical analogue 
method is still more accurate than the new digital method 
(except for the last row, commented below). 
In table 2, the results of all the applied methods are put 
together. The information provided is: Æ a, b — catenary 
parameters, m, — the mean squared error of k. F, mj, — the 
horizontal tension strength of the rope component and its error, 
calculated by the law of error propagation based on equation 2 
(weight unit of the rope q is considered as errorless). 
During the experiments, other line following algorithms were 
tested; however, the one which gave the lowest m, is the only 
one presented here (paragraph 4), as it determines the catenary 
most accurately. 
It can be seen that the lowest m, was received by the new 
method. The value of the tension strength is the same for all 
methods with the exception of the classical photogrammetric 
method (traditional). However, 1 N difference is of no practical 
consequence. Although we do not know the real tension 
strength, the values from independent methods have proved 
their reliability. The most important thing is that the value 
received by the new method is in the range of the values 
received by other methods. 
  
  
  
  
  
  
  
Method applied k[m] | a[m] | b[m] my | Fo [IN] my 
[cm] [N] 
geodetic survey based | 51.80 | 49.46 | -67.55 | 3.8 | 457.0 | 0.3 
method 
classical 51.94 [49.63 | ~67.81 | 6.9 | 458.0 | 0.6 
photogrammetric 
(traditional) 
classical 51.82 | 49.49 | -67.59 | 42 | 457.0 0.4 
photogrammetric : 
  
  
  
  
  
  
  
  
  
  
     
   
  
  
   
    
   
   
  
  
  
  
    
  
  
  
   
  
  
   
   
  
  
  
  
  
  
  
Angle Image Terrain c; mm 
Camera name precision pixel size | pixel size Beale 
» factor 
um mm 
3.0 
Photheo (assumed 
19/1318 32 for 0.7 230 194,92 
(traditional) analogue 
pictures) 
Photheo 
19/1318 22 21.0 4.8 230 | 194,92 
(scanned) 
Kodak DCs 372 9.0 8.1 900 | 50.00 
760 
Kodak DCS 
760 (sub 0.5 0.12 1.1 900 50.00 
pixel) 
  
  
      
     
    
Table 1. Accuracy characteristics of pictures. 
Table 2. Comparison of obtained results. 
It has been said above that the line following algorithm results 
in sub pixel resolution (paragraph 4). A very simple analysis 
was carried out to determine the sub pixel resolution of the line 
following algorithm. The mean square error of the rope position 
determined by the residuals was 1.1 mm, with the assumed 
scale factor of m, = 900. In the image we have 1.1 um, which 
results in 1.1/9 = 0.12 pixel (9 um is a pixel size). 
Table 1 can be slightly modified. The new method precision for 
sub pixel resolution is put in the last row of the table. The result 
from the last row should not be paid too much attention to, as it 
is merely an illustration. Pictures in different scales will 
probably result in slightly different sub-pixel resolution. More 
experiments on different test fields and with different scales 
should be done in order to draw more reliable conclusions. 
5. CONCLUSIONS 
A system for the semi-automatic determination of the 
tension strength of a stay-rope has been elaborated and practical 
experiments have been carried out. The system is based on the 
photogrammetric method, which has been adopted in new 
conditions. New technological processing has also been 
developed. This requires special conditions to be satisfied for 
setting out the control points (paragraph 2.2). This also requires 
high tech electronic equipment such as a laptop, digital camera 
(Kodak DCS 760 for the described project), and a picture 
transfer between camera and computer (cable connection). An 
elaborated computer program is the heart of the system. It 
enables such operations as image visualization, control points 
measurement, showing the placement of the ropes in the image, 
line following and carrying out the computing of the tension 
  
   
Internatioi 
strength. 
operator. 
measure 
processing 
Some furtl 
listed in pc 
£y" FEMA NLX£VFAI) 
4 f^ 
"-— ih AN FAS 
^^ 
AS 
if Z. Ka 
platfc 
Archi 
2000. 
[2] Hans 
gener 
Archi 
XXX 
[3] Armi 
and : 
Journ 
numb 
[4] Karl 
Dümi 
[5] Paul 
image 
I i I 
posre 
odcia 
Zeszy