109 
Fig. 4 - D.T.M. automatically correlated with 
manual integrations (low altitude aerial coverage). 
Fig. 5 - D.T.M. automatically correlated with 
manual integrations (medium altitude aerial 
coverage). 
operator, who must work ¡ike a photointerpreter of 
extreme competence. The possibilities offered with 
the Digital Photogram metric Workstations to 
observe the stereoscopic modei by more operators 
and/or researchers becomes essential in the 
interpretation of the objects reading and it will be 
more and more exploited in future. The vector 
restitution of the large area of Cerveteri ruins has 
been realised with analytical classical procedures, 
after a small aerotriangulation block of 8 models. 
3. Digital ortophoto of large and very large 
scale photographs 
The more interesting experience has been 
performed by executing a mosaic of orthophotos 
for the large scale images, in a rather wide area of 
the excavations, and a single orthophoto of the 
area that has shown successive modifications of 
the excavation for a very less wide surface (around 
15,000 mq). in both cases a regular grid D.T.M. 
was made the area of interest, excluding some 
belts with presence of thick vegetation, in which it 
would be difficult to obtain a coefficient of 
correlation of satisfactory value, 
In the first case the grid was of 10x10 m, resulting 
the step adequate to the texture and to the shapes 
Fig. 6 - Orthophoto with overiayed D.T.M. (grid 
5x5 m, low altitude aerial coverage). 
present in the medium-scale images, for most of 
the collimated points. A manual integration of the 
D.T.M. has been necessary only in small parts of 
images, where the presence of trees and elements 
with different perspectives has needed the 
presence of the operator. 
In the second case, although the step was denser 
(5x5 m), the automatic correlation has failed, like it 
was foreseeable, in the areas with sudden 
gradients and with dense vegetation, that they 
have prevented a good collimation (also manual) 
because of different perspective vision, being the 
aerial coverage realised with a normal overlap (60- 
70%). In fact, for being very low the relative height, 
the objects (in the interested excavation area) with 
little variations of elevation they have been easily 
recognized in shape and texture; instead the 
vegetation, above ail those with large hair, they 
have generated the presence of some holes in the 
D.T.M., that is the absence of points where there 
are those objects (trees and so on). 
In synthesis, with the presence of strong 
perspective differences both in automatic and 
manual correlation an intensification of the grid 
doesn’t furnish appreciable improvements. 
In the following elaboration of the orthophotos, this 
problem has generated the stretching of a small 
number of interpolation mashes in D.T.M. creation, 
with the consequent deformation of some parts of 
the images. In these cases, the best correct thing 
to do it would be to ignore these small zones that 
distort the total vision of the object. 
4. Conclusions 
The experiences exposed in this paper, executed 
as a test for the Institute for the Etrusco-ltalic 
Archaeology of the National Research Council in 
Rome, they have allowed us to know some of the 
problems related to archaeological applications 
that are of not simple interpretation and