469 
try triangle 
Journal of 
eight mod- 
^stem. In: 
roceedings, 
-72. ISBN 
n using cir- 
tern Analy- 
1077. 
H. Rush- 
:dings, An- 
dison Wes- 
W„ 2000. 
0, 2000). 
, K., 2002. 
itation sys- 
ision 2002 
n, Graz. 
and Gool, 
ritage from 
;ium, Pots- 
»hotogram- 
ern Recog- 
: compres- 
;actions on 
The Vir- 
Inter- 
72-1:1997. 
s/vrml97. 
t and com- 
indamental 
sion 24(3), 
Berns, C., 
est heroon 
DIGITAL SURFACE MODELS FOR ARCHITECTURAL HERITAGE ANALYSIS. THE SURVEY OF THE 
AQUEDUCT OF “LOS MILAGROS” IN MÉRIDA, SPAIN. 
Fernández Martín JJ.; Martínez Rubio J; San José Alonso J 
Valladolid University, Spain. 
School of Architecture 
Laboratory of Architectural Photogrammetry. 
WG 6 
KEY WORDS: Architecture, photogrammetric recording, digital surface modelling, visualization techniques, 
monument surveying. 
ABSTRACT: 
The documentation of a site such as a Roman aqueduct offers an excellent opportunity for testing the application of 
software resources and techniques which are usually applied on civil works and modern cartography. The linearity of 
this monument allows a very structured and systematic information management. We present a graphic documentary 
work which is part of a bigger project aimed at the consolidation, restoring and illumination of the aqueduct. This work 
serves as a basis for all of the involved sub-studies regarding lithology, archaeological analysis, rock and fabrics tests, 
and structural analysis. It also acts as the graphic support for thematic maps related to biological colonisation, vandalic 
attacks (graffiti) and other pathologies, history of restorations, etc. 
The basic support for the survey works was based on widely-experienced topographic and photogrammetric 
techniques; mainly the use of robotic reflectorless total stations, analytic photogrammetric instruments (Leica SD2000 
and Adam MPS2), digital systems,‘rectification etc. All of these techniques are all well-known among our community, 
but in this particular case further processes that imply some innovation in the field of Heritage Recording have been 
our contribution. 
Triangular Irregular Networks (TINs) offer very interesting possibilities for deformation analysing, exact volume / 
mass calculations, material loss evaluations, measuring degrees of metheorissation, real time sectioning, automated 
slope field graphing, etc. The digital modelling of the building surfaces represents an important qualitative approach to 
overcome the limitations of wiremeshed models, as those digital models present a true three-dimensional topology that 
makes the interpretation of the stone limits truly objective. 
Beyond the application of the above mentioned quantitative analysis, surface modelling offers new ways to get realistic 
virtual renderings. Three dimensional views that reveal qualitative aspects, such as texture and the response to different 
illumination ways, have been other goals of the obtained results. 
1. INTRODUCTION 
This paper aims at the description of a practical application of 
some tools offered by computer graphics, which are specific of 
the field of civil engineering in the context of the analysis of 
architectonic heritage. Professionals working on the 
preservation and documentation of heritage recordings very 
often have to face the problem of the inefficiency of graphic 
design programmes, when they try to present and manipulate 
the documented objects. In quite a few occasions, something as 
simple as achieving partial visualizations, without the presence 
of any obstacle or occlusion, implies certain problems that 
make it necessary to make use of strategies which are 
desproportionally complex in relation to the problem itself ( big 
fragmentation in partial files, complex structures of layers or 
level with the only aim of differentiating orientations...). This 
is even more reproachable when, sometimes, this difficulty 
makes it necessary to do without some useful information, in 
order to make the visualization more legible. 
The problem of spatial superposition can only be satisfactorily 
solved by the application of some graphic tools which are not 
currently found in the most frequently used CAD programmes. 
In any case, the person doing the drawings must keep in mind 
that the line-strings he is drafting are, in fact, sides of objects. 
That is to say, they will have to build topological structures. 
However, as we all know, when doing restitution one tends to 
create very complex drawings which look like wire tangles. 
Figure 1. 3-D files of the Aqueduct.