GeoTeX: A GENERAL POINT DETERMINATION SYSTEM
I.Colomina, J.Navarro and A.Térmens
Institut Cartogràfic de Catalunya
Spain
ABSTRACT:
The GeoTeX system of the Institut Cartogràfic de Catalunya (ICC) is a general geodetic and photogrammetric
point determination system which is able to deal with any type of geometric functional model. GeoTeX is
suited for research as well as for production purposes and can be easily extended to incorporate new models.
Examples of tasks that the system can master are: spatial triangulation —SPOT images—, aerial triangulation
with kinematic GPS derived positions, conventional geodetic network adjustment, DTM surface information
and combinations thereof. In the paper, two components of the system are described: the discrete model kernel
based on discrete mathematical techniques and the I/O kernel based on a general formal geodetic model and
its associated formal grammar based interface language.
KEY WORDS: abstract data types, combined adjustment, compiler-compilers, data standards, grammars,
network discrete models.
1 INTRODUCTION
In a sense, network point determination is an old-
fashioned topic in photogrammetry and geodesy. It
can be stated at least that the subject is no longer
appealing to the academic research community.
On the other hand, fast, accurate and reliable point
determination is still a must for private companies
and official agencies in the allied fields of geodetic sur-
veying, photogrammetry and remote sensing. These
organizations face a variety of problems ranging from
huge data sets, to changing hardware/software envi-
ronments, to the extension of the traditionally sup-
ported observable types (for instance the testing of
new mathematical models or the use of new sensor
types). As a consequence, remarkable efforts are still
being made mainly by instrument /software manufac-
turers and by research/development departments in
land surveying agencies in order to improve and keep
the software systems for point determination up to
date.
Two additional introductory remarks are in order
here. First of all, the principles and concepts estab-
lished by the research activities related to the for-
mer ISPRS Working Group III.1 (period: 1984-1988;
subject: accuracy aspects of combined point deter-
mination) have not been transferred to practice with
the exception of industrial photogrammetry and GPS
aided aerial triangulation. Secondly, advantage of
already existing progress in computer hardware and
software must not only be taken to speed up the old
software but also to develop superior methods and
procedures.
At the ICC, besides the above general remarks,
there are particular motivations behind the research
and developments reported in this paper. À unique
point determination system both suited for research
and production was needed since the development
656
and maintenance of two different systems cannot be
afforded. Similarly, it is not cost feasible to develop
and maintain a system for geodetic networks, a sys-
tem for aerial triangulation and a system for spatial
triangulation.
The authors claim that this kind of a contradic-
tion between research/production software and tai-
lored/general software can be overcome if proper
mathematical and computer science tools are used.
Another claim is that the benefits of this global ap-
proach are not solely of interest to small groups. They
are: a cost reduction in software development, ac-
quisition and maintenance; a closer collaboration be-
tween teams traditionally involved —as well as tra-
ditionally separated— in point determination tasks;
and the introduction of factors of rationality and
coherence in the corresponding point determination
projects.
2 ON GENERALITY, ADAPTABILITY
AND PORTABILITY
In the context of point determination systems, gen-
erality is the power of mastering a broad range of
[network] observations, conditions and parameters;
adaptability is the power to incorporate new ones;
and portability, as everywhere else, is the power to
run on different platforms with minor or no changes,
To say that generality, adaptability and portability
are desirable properties of a software system is com-
mon place. Rather than restating this, this section
aims at describing a specific example of how the pur-
suing of those properties has lead to certain design
concepts.
fe
né
re
pi
pe
fe
sic
CO
og
du
tr)
ge
sti
Sul
rea
ple
tio
mu
ser
def
prc
vat
DI
2.2
Sor
exh
men
prof