s not correspond 
t define in detail 
of the model has 
e field and the 
imark, anchored 
| the model can 
mations of each 
ienting all the 
ditional survey, 
ther words once 
eristics of an 
ir surface or a 
cessary to mark 
sections. In the 
ectangular trend 
tions) and in the 
end. 
| all its parts of 
ks or contacts 
) grid of about 
, with a final 
operation was 
1e object in its 
losi” and the 
covering of the 
ity of points’ 
photographs in 
dn't manage to 
eoscopy of the 
being the 
  
Figure 6. Dolium fragment. 
3.3 Procedure issues 
photographs shot almost exclusively with vertical axis and with 
continuous trail capture, we got an angle with too small 
incidence on the screening radius' on the single point shouted 
by two photos in succession, therefore not suitable to a 
correlation with Photomodeler, beyond the fact that all the 
points of vertical walls are of difficult determination. 
4. ARCHEOLOGICAL DATA WEB ACCESS 
We plan to offer the collected data to the community, 
(archaeologist, photogrammetrist, or every body interested), 
that is why we decided to make a web presentation allowing the 
exportation of photogrammetric data as well as archaeological 
ones. 
4.1 Uniform data 
In the case of the second site (Relitto dei Marmi) we had to 
survey and document a cargo of 11 white marble blocks and 2 
shaft of column. There the survey was here the way to get 
dimension and position of each pre-identified piece of marble. 
A survey was made some years ago, manually, by Sergion 
Bargagliotti [Bargagliotti, 98], and his report give a planimetric 
map and dimension and approximate weight for these elements. 
The photogrammetric survey gives us, firstly a good 
confirmation of the initial result, a 3D representation of the 
entire site (in VRML) and a way to publish the result using the 
3D representation as an interface to the archaeological data. All 
links on an orthophoto and the same links available through a 
VRML representation of the site. Under these conditions, rather 
easy, we can offer a bi-directional link (from graphical 
representation to data or from data to its graphical 
representation). 
All the data are accessible on the web [Cipa-uwp, 2002], by 
some links on an orthophoto and the same links available 
through a VRML representation of the site. Under these 
<?xml version="1.0'?> 
<amphora> 
<generalAmphoraDescription 
subClass-"Dolia-Meloria-D" 
type-"Panse" 
name="Panse-123" 
localisation="surface" 
nbFrgmt="1" 
> 
Figure 7. Shows a fragment of XML text, it describes a Dolia 
  
jhwe Bide de $ pow 
d] ewe | cdrom? Hemente LA 
  
Figure 8. screen shot of the vrml with links to specific data. 
conditions, rather easy, we can offer a bi-directional link (from 
graphical representation to data or from data to its graphical 
representation). 
4.2 Heterogeneous data 
For the first wreck (Relitto dei dolia) the cargo was composed 
by dolia, and now all partially destroyed. The survey and 
documentation of this cargo is more complicated. 
In this case archaeological documentation is more complex. 
Data to manage are mainly heterogeneous : 
Data heterogeneity comes, in part, from the differences between 
studied dolia and parts of dolia or also between dolia 
themselves. Several type of data have to manage in the same 
way, recorded in the same file, without any loss. On the other 
hand we have to mix measured data as well as archaeological 
data coming from human experts. 
XML is the standard for federation and exchange of data 
between heterogeneous applications. It's a way to formalize 
data in which the document structure is described in a separate 
file that allows the user to seek in the document and get only the 
relevant information. 
XML! is a recommendation of the World Wide Web 
consortium (W3C). XML language is designed to associate a 
tree structure to data. The tree structure is expressed in the data 
thanks to tags (an opening and a closing tag defines an 
element). A set of attributes can be associated to each element. 
The content of an element can be either empty of composed of 
an ordered list of elements and texts. [Drap, Bruno, Long, 
Durand, Grussenmeyer, 2002]. 
5. CONCLUSION AND FUTURE WORK 
This work started one year ago and is the result of a really 
interdisciplinary approach and a will of dissemination of 
archaeological knowledge. We started this common work by 
sharing between us vocabulary, knowledge and experience. In 
this way we present this work in the framework of CIPA 
underwater photogrammetry work group, in order to share 
experience, through the web, on surveying and documenting 
archaeological excavation. Currently we have done the first 
experimentation on the (Relitto dei Marmi), the cargo of e 
marble blocks. This work is visible on the web site [Cipa-uwp, 
2002]. 
We showed that surveying and documenting a shallow 
underwater excavation was possible even with few expenses, 
one of the goal was to develop tools and methods easy to use 
for archaeologists. 
—33]—