C1PA 2003 XIX th International Symposium, 30 September - 04 October, 2003, Antalya, Turkey 
538 
2.1 Topographic survey 
An appropriate network has been applied in all cases of the 
study; vertices and topology have been projected to guarantee 
the collimation of all necessary points to describe form and 
geometry of the structure. Survey of details is an onerous phase 
in old buildings: the problem of measuring is moved to criteria 
of point selection. In the process of measurement, the elements 
deriving from it have enabled a location of the points and the lines 
in the space. The approximation grade is linked with a significance 
and a quality of the selected points. Very often a design attenuates 
or accentuates the geometry leading to formulation of the erroneus 
interpretation: non-respected alignments, forced or non-measured 
symmetries. 
A total station, without reflecting prism, has been the instrument 
most used used in this kind of survey. Free from the requirement 
to signal the points to be located, it became possible to concentrate 
on the selection of the same; this imposes an accurate plan of 
measurement of a detail. In such a way it avoids the risk of not 
having sufficient data to describe the shapes or, on the contrary, to 
have an amount of information requiring costly and patient work 
on simplification of the graphic design. During the recognition 
phase, drafts have been issued with points, duly numbered and 
linked among them; these points should lead to a signifcant line. 
In the subsequent acquisition stage, the possibility of using an 
identification code allowed the pre-management of the data 
into files that group the families of the points and that have 
been memorized at different layers in order to facilitate the 
management in cad software. File management is fundamental: 
a subject is described by means of the points among which the 
linear interpolation is programmed; a descriptive structure that 
derives from there defines wireframe lines, so an operator decides 
what should be in-between the lines. The same lines are, therefore, 
a result of the choices aimed at the creation of a model that, as 
always, is a simplification and so a scheme of reality, furnished by 
a sufficient descriptive accuracy in order to respond to the research 
objectives. It is advisable to work with a very clear project with 
regard to a data bank that, step by step, is being implemented. In an 
tree shaped organisation, for example, it is possible to distinguish 
the lines of different nature: structural, wall apparatus, the profiles 
of ornaments and decorations etc..., changing the rules, already 
consolidated ones, of photogrammetric acquisition from the 
observation of the model to observation of the reality. 
2.2 Photogrammetric survey 
As well-known, photogrammetry has numerous positive attributes 
in architectural recording, however there are some limitation of 
the photogrammetric product that digital techniques allow us to 
overcome. In any case a vast volume of primary data is quickly 
captured and the problem of chosing what kind of data, geometric 
or thematic, has to be plotted, is posticipated and can be discussed 
in the laboratory in accordance with the users survey. Digital image 
processing has simplified the pratical approach to photogrammetric 
survey in such a way that even non experts in photogrammetry can 
perform the plotting and interpretation phases. So, while geometric 
problems can be understood by photogrammetrist, the other aspects 
of the survey can be performed by conservation specialists. 
With digital techniques it is possible to distinguish acquisition 
unit (images, coordinates to orient acquired images, measure and 
computation to estimate orientation parameters), from restitution 
unit (finding interesting points, classifying information in logical 
levels representing results in an efficient way). All the topics of the 
acquisition unit requires a specific technical expertise and can be 
correctly designed and executed by the photogrammetrist. 
In the same way all the topics of restitution units can be performed 
by the operators who can study other heritage aspects indipendently 
from the acquisition and orientation phases. 
Another simplification has been introduced with a return to direct 
photogrammetry in which the absence of control points allows 
to face up to emergency survey where the necessity to collect on 
the field, as much information as possible, as quickly as possible, 
is often conflicting with high quality metric results. We present 
one example, as work in progress, of this direct photogrammetry 
application. 
2.3 Image rectification 
To the necessary critical selection that makes a measurement and 
the graphic design an instrument of immense use have also been 
added the valuable characteristics of a photo. A passage from 
analogical to digital allows, nowadays, the more flexible use of 
a photo imaging that releases a complete and “global memory” 
in camparison to analytical designs achieved by topographic 
measurement. 
Aiming at a reduction of field work, and in order to document in 
the best possible way the metric difference of the surfaces, digital 
rectification can be used thanks to which it is possible to define in 
detail the values achieved outlining the areas where the contours 
have been measured by topographic or photogrammetric use of 
points. A rectified image can be used to integrate the other kind 
of information either in evaluation of the surface conservation 
state or in 3D model texturing deriving from photogrammetric 
restitution. 
We present one example of rectified image use to furnish a 
base to analyze the wall texture of Povil Casaforte (Aosta) in a 
stratigraphic study of the facades. 
3. OUTPUTS 
The most commonly applied products in cultural heritage 
continues to be the line drawing produced from topographic or 
photogrammetric plotter, normally at 1:20 or 1:50 scale. In these 
products we have to pay attention to the quality of the drawing; 
to obtain a useful drawing for conservation requests is demanding 
for the operator and can be achieved in practice by working with 
large scales. A key change in working methods has taken place 
relating to the digital capture of data, making the outputs of the 
photogrammetric process much more flexible. Another implication 
of this change is that output as three dimensional model is now 
commonplace. 
The rectified photography method continues to provide a valued 
product as a base for a thematic survey or as a texture for 3D 
model while true digital orthophoto (Boccardo et al., 2001) is 
becoming a powerful instrument of representation in architectural 
application thanks to a new approach which uses several images 
and a DDEM (Dense DEM) of the object in order to produce 
an orthogonal projection of complex objects. It is necessary 
to remember as in the heritage field, photography (image) is a 
record of great significance. At the very large scale extracting 
information is problematic and the object complexity requires 
personal interpretation. A new digital product, a stereo-photomap 
(Dequal et al., 1999) is now available to explore direclty the 
photogrammetric model. All the images of a block, previously 
oriented, can be explored as a unique large stereomedel. This new 
instrument allows the conservator “to navigate” around a building 
without any need of a preliminary restitution: they can select and 
measure directly only what is interesting for their study. Just as 
with the long history of photo-interpretation, direct stereoscopic