C1PA 2003 XIX"' International Symposium, 30 September-04 October, 2003, Antalya, Turkey 
271 
- A monument, located at Praia Vermelha, Rio de 
Janeiro, 2000, (PhotoModeler), see (2, Barbosa) 
- Rock Art Photogrammetry - The Bear at Fonte 
Grande II Canyon - Uibai, Bahia, Brazil, CIPA 
2001,(PhotoModeler), see (10, Gilson D. Koatz) 
3.2.2 Completion of Simplified Ground Control 
However, in the case of large objects, one needs to 
complete the simplified ground control by densification. 
This can be done in different ways: 
- by processing with software like ORPHEUS and 
ORIENT of the Institute of Photogrammetry and Remote 
Sensing , Vienna University of Technology, see (4, 
Gomes: The Empress Manor-Project). An orthophoto of 
the façade was produced by IDL-Software. 
- by photogrammetric measurements; 
Almost simultaneously to processing with ORPHEUS and 
ORIENT in Vienna, the same stereopair of the 
Empress Manor -Project with incomplete (poor) ground 
control was introduced in the AVIOLYT BC 2 of 
AEROFOTO CRUZEIRO S/A. In the relatively 
orientated model about 33 points were measured. 
Later on, the single model was adjusted by the 
program PAT -M. Author: Hanns J.C. von Studnitz, 
AEROFOTO CRUZEIRO S/A, Rio de Janeiro. 
A digital restitution of the same façade was executed in 
the Digital Video Plotter-DVP of IME. 
- by additional topographic measurements. 
In this case, one needs a coordinate system outside the 
object system, the so-called 
3.2.3 Arbitrary local coordinate system 
Origin : 
X = 1000.000 
Y = 500.000 
Z= 100.000 
Figure 2 - Definition of the two coordinate systems 
3.3 Topographic Measurements 
3.3.1 Polar measurements by use of a Total Station 
Initially it was planned to use a Total Station LEICA TCR 
307 in reflector-less mode with simultaneous 
trigonometric levelling from a baseline of two stations in 
order to determinate polar coordinates of control points. 
But it turned out that due to very oblique pointings, the 
return signal to EDM was too weak and thus the distance 
measurements failed in about of 40 % of instances. 
Therefore we decided to use the intersection method for 
all points of the project and to consider the polar 
coordinates for comparison purposes only. 
3.3.2 Forward intersection 
The spatial coordinates to a total of 42 points (mostly 
targets) were determined by forward intersection in the x- 
y-plane with simultaneous trigonometric levelling. In the 
beginning from two stations in the arbitrary system, in 
later stages from three stations directly in the object 
system (see 3.6.2). 
For safety reasons we strongly recommend a double 
forward intersection using two baselines, for example B- 
A and B-C, (see figure 3). 
Figure 3 - Double Forward Intersection 
3.3.3 Comparison between polar and intersection 
method 
Coordinate differences for 15 determinations were 
compared and showed the following standard deviations: 
± 6.61 mm in x-coordinates (1 point out of 
tolerance) 
± 8.26 mm in y-coordinates (2 points out of 
tolerance) 
3.4 Computations 
The spatial coordinates ( x,y,z ) were computed by use of 
a program developed based on manual calculations for 
plane forward intersection combined with trigonometric 
levelling by the author Walter da Silva Prado, (see 
fig. 4).