)TTERS 
uneiro 
iue to 
Their 
Mr, by 
block 
and 9 
nates 
using 
iques 
s les 
pour 
n les 
dèles 
photo 
ceaux 
die 
ndern 
Die 
r die 
en: 
ngige 
und 
und 
g on 
"rol, 
utilizing different observations, searching for small blunders 
and the like. The relevant photogrammetric measurements are 
nearly always performed with a comparator (mono or stereo), 
providing a least count reading of 1 um. 
Yet, in photogrammetric practice there are relatively few 
comparators in use. Even with the ever increasing emphasis on 
digital systems, the analogue stereoplotter is still the most 
popular photogrammetric instrument. This fact is even more 
pronounced in third world countries, where the analogue stereo 
plotter is likely to remain the photogrammetric  workhorse 
because of its versatility and relative insensitivity to 
variations in electrical power and related changes in the 
operational environment (e.g., breakdown in air conditioning 
systems, etc.). 
The majority of photocontrol for medium and small scale mapping 
is being determined with independent model triangulation using 
analogue plotters. 
In this study, the use of the analogue plotter for aerotriangu- 
lation was investigated in a more general fashion. 
ANALOGUE PLOTTER MFASUREMENTS AND THEIR EVALUATION 
Commonly, model coordinates provide the data for independent 
model block adjustments. There are, however, other possible 
approaches as well, namely the utilization of the plotter as 
stereo- or mono-comparator Levy, 1964, Boniface, 1969]. 
-— model coordinate measurements 
After performing an optical-mechanical relative orientation, 
x,y,z-model coordinates are measured in a local model 
coordinate system. These are normally input directly into an 
independent model block adjustment which may have additional 
parameters. However, it is also possible to decompose them 
into photocoordinates for processing in a bundle adjustment 
using reseau or fiducial coordinates fJeyapalon, 1972]. A more 
general method is to record the orientation elements of both 
projectors which slightly increases the observation time. Then 
the rotational matrices can be formed for both projectors, and 
with the collinearity equations photo-coordinates are derived. 
Since each bundle is part of two models, the two respective 
halves have to be joined by a similarity transformation via 
common points. 
-- model measurements in stereo-comparator mode 
In addition to x,y photo-coordinates, x- and y-parallaxes are 
measured in a stereo-comparator. Since an analogue plotter 
does not have direct parallax-measuring facilities, one 
specific motion each is used to eliminate the respective 
parallax. The reading thus obtained is easily transformed into 
the parallax if needed. The x-parallax is replaced by the 
z-model coordinate, while the y-parallax is best measured with 
b., and if a b_-motion is not available, with © (Harley, 1971]. 
Thus, instead’ of performing a relative orientation on the 
plotter, the y-parallax is eliminated - and measured - at each 
point thus reducing the observation time by at least one third. 
Then an analytical relative orientation is performed and model