ased 3D object repre- 
Computer 20(8), 19— 
M. J. (1983), 3DPO: 
tion system, in ‘Proc. 
onf. on Artificial Intel- 
.. 1116-1120. 
C. R. (1992), Aspect 
ications in digital pho- 
ongress 1992, IAPRS 
33—645. 
D-based computer vi- 
ational graphs’, /EEE 
Machine Intelligence 
ognition by computer, 
IT Press, Cambridge, 
> photogrammetry — 
in ‘Proc. Comm. V 
; and Machine Vision, 
135. 
), CAGD-based com- 
tern Analysis and Ma- 
1193. 
'a Representation and 
] Fundamental Princi- 
1a 1996 
Long Terme Photogrammetric Measuremend of Rockfill Dam 
Jozef Cernansky 
Slovak Technical University Bratislava , Slovakia 
ISPRS Commission V , Working Group 3 
KEY WORDS :Bundle, Adjustment, Analytical, Monitoring, Network, Comparison, Parameters, Accuracy 
ABSTRACT 
Two photogrammetric methods the SAS method and ORIENT software were used for the photogrammetric measurement of the 
rockfill dam. The results of these both methods were compared with a ground control survey in one epoch of the measurement. The 
stability of the monitoring network was established by concrete pillars with forced centering for the theodolit, terrestrial camera 
and targets. The required accuracy has been kept in the range of 1-2" for the angle measurement and 1-2 um for the 
photocoordinate measurement. The results show that the achieved accuracy of 2-3 mm in spatial displacement of a rockfill dam 
determinated by the SAS method and by ORIENT software were comparable to the ground survey. 
1. Introduction 
Since 1977 we have determined the spatial displacements of 
the rockfill dam Bukovec (East Slovakia) using the analytical 
photogrammetric method called: Separate Analytical Solution 
method - SAS method. This method was presented at the 
Congress ISPRS in Rio de Janeiro 1984 and at the Conference 
Optical 3D-Measurement Techniques III in Viena 1995, 
(Cernansky, J., 1995). 
The SAS method has been applied in 23 epochs of the 
photogrammetric displacement measurements carried 1 - 2 
times per year. The method use the horizontal and vertical 
angles measured by a theodolit with forced centering directly 
on the photogrammetric stations. The coordinates of the 
projection centers were determined by ground survey with the 
eccentricity of the camera entrance pupil and its bearing. 
For the spatial displacements of that rockfill dam was utilising 
also the ORIENT analytical adjustment software (ORIENT, 
1991). It was possible to compare the spatial coordinates 
determined by the SAS method and by the ORIENT with 
ground survey in 1994. 
2. Determination of the spatial coordinates by analytical 
intersection using the SAS method. 
For the displacement measurements of the dam, a monitoring 
network with reference, control and detail (observed) points 
was established (Fig. 1). The reference points were situated 
around the body of the dam and served as ground survey — 
theodolite stations. Some of them were the camera stations. 
The detail points were placed directly on the body of the dam. 
All these points were constructed as concrete pillars. The 
ground survey coordinate system XY is visible from Fig. 1, Z- 
axis directed to the zenith. 
Y 
F15A 105 
  
  
Fig.1 Monitoring network of the dam 
Ground survey 
The ground survey angles of reference points, control and 
detail one were measured by precise theodolit Wild T3 in three 
sets with an achieved accuracy: 
ma=0,7" mp=1,1" (1) 
All those points were marked by the special circular targets 
both for the ground survey and photography. The camera 
stations F1, F3, F5 (Fig. 1) are identical to the reference points 
of monitoring network. 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B5. Vienna 1996