04
he
he
ot
he
IS
of
y,
he
a
he
„a
ng
on
he
la
nd
for
bs,
ne
ers
se
ult
by
m
ed.
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B6. Istanbul 2004
plotting, because would allow to locate with more ease objects
featuring a contrasting colour, such as roofs, trees, green area,
water surfaces. A realistic possibility would be to use HRSI
pan-sharpened images. Pan-sharpened images are generated by
merging the colour information contained in the lower
resolution visible (or visible/infrared) multispectral bands with
the geometrical information contained in the higher resolution
panchromatic band. The result of the processing is a natural (or
false color) pan-sharpened image, with the resolution of the
panchromatic spectral band.
It should be noticed that reported vector data directly come
from interpretation of the image and have not been edited. In an
operational production process, also this stage would be carried
out, involving recovering of orthogonality, parallelism an the
like.
In the test area, other two cartographic product were available,
i.e. a colour orthophotomap at 10,000 scale and a portion of the
regional raster map (CTR) at the same scale. Because these
maps are geocoded into the Italian official cartographic system
(Datum “Roma40” based on the Hayford ellipsoid) using a
modified Gauss projection (termed as “Gauss-Boaga”), a
transformation to UTM-WGS84 system used for the IKONOS
orthoimage has been computed. Due to the limitation of the
area, a 2D conformal transformation would result enough
accurate.
Vector layers extracted from IKONOS orthoimage have been
overlayed to both orthophotomap (Figure 6) and raster map
(Figure 7). As can be easily understood, roads have been
extracted with a high accuracy, completely conformal to
tollerances adopted for 1:10,000 maps. In Italy, we usually
adopt a cartographic planimetric tolerance which is twofold the
size of the map resolution; in case of 1:10,000 maps, this
tolerance adds up to 4 m.
Of high accuracy and completness is the drawing of building as
well, in particular when compared to the orthophoto.
Comparison with respect to raster map shows several
differences in geometric positions of buildings, but this fact is
probably not due to the quality of the IKONOS orthoimage.
Figure 4 — Patch of the orthoimage derived from an IKONOS
over tha area of Lecco used in the test for evaluting the
possibility for information capture from HRSI.
Figure 5 — Overlapping of the extracted vector layers to the
IKONOS orthoimage patch.
Figure 6 — Orthophotomap (1:10,000 scale) derived from aerial
photogrammetry with superimposed vector layers extracted
from IKONOS orthoimage.
Figure 7 — Raster regional map (CTR at 1:10,000 scale) with
superimposed vector layers extracted from IKONOS
orthoimage.