Archaeological Survey: Experimentation and Direct Observation 
183 
Regarding the second test, the final result was a more exact graphic model of some arcades through a greater number of points for 
single photogram. The model uses the same number of control points but a greater number of referenced points for the definition of 
the various structural planes. 
5. ISSUES AND RESULTS (M. Canciani, G. Spadafora) 
The main problems during the test of the survey method concerned the quickness of the procedures and the accuracy level of the 
measurements, not always equal to the appointed task. Indeed some procedures could be improved in effectiveness and reliability, 
for instance by using an electronic theodolite that automatically generates a three-dimensional DXF file, or by setting up pre-printed 
markers to point out the control points more distinctly. Concerning the acquisition of the images, the procedure could be improved 
by using a digital camera that also makes easier the calibration process. Besides, it could be useful to go deep into the check of the 
data as well as into the accuracy of the geometric model. By the way, the use of new and updated instrumentations would involve 
obviously an increase of the survey costs, but would reduce the time required for the computer elaboration and would improve the 
accuracy and the reliability of the process. 
As regards the obtained results, the more convincing part of the work seems to be the definition of a navigable vrml model 
exportable in other cad programs, and definable "opened model " for following elaborations if by computer (any kind of horizontal 
or vertical section or axonometry) or by hands (drawings executed on the orthogonal projections cad printed). 
The program therefore allows a remarkable saving of time in the realisation of three-dimensional schematic models in which the 
interest is more in the general architectural appearances than in the details, with the advantage to have measurable objects in 
orthogonal projections in short time. In this sense Photomodeler can be counted among the expeditious survey methods whereas with 
this denomination we mean the possibility of drawing with such a tolerable deviation in the final scale of graphic representation. 
One of the advantages of Photomodeler use is the possibility of obtaining a view of the elevation in orthogonal projection with 
rectified photographs, from a three-dimensional object; vice versa the rectifying procedure and the following assembly of single 
photograms in the drawing needs longer operating time and produces a result limited to the single specific projection. 
The final result can’t only end with the elaboration of the numerical model. That, in our opinion, should be the statement for 
following analyses that graphically synthesize the gathered data and turn them into signs able to communicate the nature of the 
represented object unequivocally. In fact, the model of the whole eastern wall, in his approximation, can be used to serve to consider 
the structural and morphological character and it gets evaluation only if connected with its environment; vice versa the more 
accurate arcade model simplifies the drawings of structural elements. The last aim of the survey can’t, in fact, be the return of a 
mimetic model of the reality. The numerical elaboration of the single arcade was, therefore, set up in a way that following graphic 
elaborations could be readable in a scale of 1:20, suitable for taking back the characteristics of the materials and the presence of the 
plasters. 
Fig. 17: Cryptoporticus in thè Casa dell’Ancora. Eastern elevation in Photomodeler 
6. CONCLUSION (D. Maestri) 
The creation of data processing three-dimensional models as well as the three-dimensional dynamic visualizations and the various 
rendering elaborations have since been taken as granted. 
A following step will be to increase and to examine in detail the solution ot the various problems in the architectural studies 
developing the collaboration between users and software houses (in this case the Eos System). Beyond the photographic precision 
and the reliability of the imagines in the process of surveying, the core of the question becomes the quality ot representation. This 
fact must be a fundamental prerogative of all architectural data processing. As usual, in our case, the drawing ol eidotipi had 
fundamental importance to improve the use of data processing programs and to choose the best representation scale.