C/PA 2003 XIX"' International Symposium, 30 September - 04 October, 2003, Antalya, Turkey
5. THE PROPOSED FLOWCHART
5.1 The image acquisition
As it is already stated, the flowchart of the new proposal (Figure
1) concerns the transformation in digital form of old maps that
are real artefacts parts of our “recent” cultural heritage. This ty
pe of artefacts because of their intrinsic and extrinsic properties
should be treated in a careful and delicate manner. Especially
the generation of their dfs require the development of a new
know-how based on concrete theory, targeting at the same time
to Low Cost with Optimal Response (LCOR) management
models because the worldwide possessors of old maps are not
only wealthy maplibraries and prosperous individuals but also
many institutions and collectors with less available resources.
A new know-how in this direction is required for the rational
generation of dfs from old map originals. Scanners that are used
in DM to create dfs of printed old map documents -or even
worse, of manuscript old maps- may produce severe damage on
them, which is an impermissible situation in our cartographic
heritage documentation and management. On the other hand
taking snapshots of old maps in the frame of IM (IM-dc, IM-
ac/d) photogrammetric skills are a prerequisite in order to avoid
misuses in reducing the map images properly.
For all the above reasons a careful implementation of a per-se
scheme is necessary requiring the testing of all possible image
acquisition modes concerning the transformation of old maps in
dfs.
5.2 The processing
The testing of the relevant dfs acquired in DM or IM (IM-dc,
IM-ac/d) is the processing step of the proposed scheme. Images
of different dfs of the same old map (or versions of it) are
treated in a quantitative environment by applying a variety of
comparative studies of a manifold of geometric and non-geo
metric (radiometric, etc.) features. The processing can be based
on:
■ Conventional and non-conventional transformations
giving point-wise and areal measures of dfs differen
ces, for further elaborations.
■ The cartometric analyses and the connections of the
old map dfs with modem relevant dfs, which is the
cartographic component of the overall scheme.
The latter opens a huge domain of research coupling digital
image processing of old maps with modem digital cartography
and allows a direct interaction between raster and vector carto
graphic processing.
In this processing context, it is clear that a new multidiscipli
nary area of research and implementation is emerging relating
historical cartography with image processing, photogrammetry
and cartography plus digital numerical analyses.
5.3 The LCOR concept
The final part of the new scheme proposed here, is the articu
lation of a protocol addressed to the possessors of old maps
aspiring digital outlooks. The basic concern for this protocol is
to keep a reasonable balance between a Low Cost in dfs acqui
sition and relevant processing and an Optimal Response for
outputs that meet the basics not only of the cartographic heri
tage requirements but also the interest for modem cartographic
research on old maps.
6. THE CARTOGRAPHIC INSIGHTS
Research on the digital treatment of old maps and their properly
derived images was one of the targets of the EC-DG X
CartoTech project (Livieratos and Myridis, 1999, see also, e.g.,
Guerra et al., 1999; Boutoura, 2000; Balletti and Boutoura,
2001). In order to illustrate the proposed method some
examples are given from the digital documentation of old maps.
Two 300 dpi dfs of the same old map (Laurenberg, 1656, Col
lection: Holy Mt Athos Map Library by the Hellenic National
Map Library) have been obtained in two different modes:
■ Applying the DM
■ Applying the IM-ac/d.
The two dfs have been conformally best fitted in order to elimi
nate the small rotation and scale differences without alteration
of the form of the map. From the two images a detail was
cropped (the island of Limnos). In Figure 2 the detail is shown
from the df derived by the DM and in Figure 3 the counterpart
from the df derived by the IM-ac/d.
Figure 2. A detail of Laurenberg’s map Insularum Archipelagi
Septentrionalis, 1665 (here the island of Limnos)
derived by DM scanning the original.
Figure 3. The same detail derived by IM-ac/d after a conformal
best fitting to the image of Figure 2.
Now the two dfs are ready to be compared in a one-to-one cor
respondence in order to depict graphically the overall statistical
geometric differences due the different acquisition methods ap
