EEG 
at the Institut Geographique National. Methods 
vary with the terrain: 
— simplified method with an instrument called 
"révicarte", with optical rectification using facets 
and graphical transfer using stereoscopy 
— photogrammetric method 
— simplified photographic method with an instru- 
ment called "photo-révicarte", with photographic 
rectification using facets and deferred graphical 
transfer 
— orthophotographic method, undoubtedly the 
best from a technical point of view, but not from 
an economic, which leads to the use of photo- 
graphy at a smaller scale, combined with a stereo- 
scopic study of photography at a larger scale. 
One of the problems to be solved is that of the 
centralization or decentralization of both revision 
personnel and instruments. 
Dr. Dubuisson spoke on the revision of large- 
scale maps (urban and rural areas), noting that 
the maps are outdated by the time they are pub- 
lished. Nevertheless, the "actuality", the veracity, 
of a map is just as important as its accuracy. 
Consequently, to be of use, cadastral maps have 
to be duly up-dated. The cost of the revision of a 
map five years old amounts to 1/3 of the total 
costs of mapping in the first place. In suburban 
areas 10% of the information changes in one 
year, in rural areas only 5%. 
Dr. Dequal described a three-tier hardware/soft- 
ware system being developed in Italy for up-dating 
and managing a cadastral data base (PP /V/12). 
The system employs computers and interactive 
communications terminals at local, provincial, 
and national levels. At the time, information was 
being collected from four sources of numerical 
data: cartographic sources (existing digitized 
maps), photogrammetric sources, terrestrial mea- 
surements, administrative sources (archives). The 
system now being established is centered on an 
IBM 370/168 computer in Rome. Local data bases 
centered on mini-computers will be connected 
to this computer. They, in turn, are accessible 
via decentralized terminals (graphical and digital 
consoles). An experimental station, in use at the 
Institute of Photogrammetry of the Polytechnical 
University Turin, was being exhibited in the Sci- 
entific Exhibition of the Helsinki Congress. 
Plans had been made to up-date the cadastral 
data bases once a week. 
Mrs. Fleming gave a talk on the use of Landsat 
imagery for map revision (PP /V/19). The imagery 
is used extensively in the unpopulated northern 
areas of Canada, its greatest advantage being 
that far fewer photographs are needed than with 
conventional photography. Roads show up very 
well (on band 5) as clearing through the forest, 
while water features show up well on band 6. 
Power lines are visible on winter imagery (as a 
white swath through the forest), and winter im- 
agery is useful as an aid to relief shading 
because of the low sun angle. The many lakes 
in northern Canada are used for control, i.e., for 
matching the imagery to maps in a reflecting 
projector. 
Mrs. Fleming observed that many outstanding 
results had been obtained in various fields by the 
revision of maps of Canada by using Landset 
satellite imagery: 
— determining the location of new elements 
PN cA ta pera RL 
mV 
re opis part ld 
having large dimensions (roads, power lines) 
— obtaining natural shading due to low sun angle 
— preparing photomaps at a scale of 1:1000000 
and at smaller scale 
— identifying the location of “off-shore” elements 
not formerly connected with the main land. 
Despite their poor resolution and their dubious 
geometry, the multispectral images produced by 
the satellite Landsat are most valuable because 
they cover vast areas (one image may replace up 
to 600 photographs at a scale of 1:60000) and 
because they are repeatedly produced, thus 
making it possible to solve certain problems of 
interpretation. Furthermore, photographic en- 
largements, matched locally to existing maps, are 
an effective means of localization. 
Dr. Doyle spoke of the difficulties of interpreta- 
tion encountered in map revision on the basis 
of satellite imagery particularly in rapidly devel- 
oping areas (cities, coastal areas). He cited as 
an example the Skylab imagery (S. 190A over 
Chesapeake Bay) which was good only for shore- 
lines. Some of the photographs taken at a scale 
larger than 1:1000000 (S. 190B, focal length 460 
mm) were of minimum quality for revision, being 
adequate for rural areas. At the time, further 
experiments in revision were being made in 
Columbia. 
In the future system of Spacelab, planned for 
1980, cartographic cameras with varying local 
lengths (300 or 600 mm) will be put on board. 
Likewise, stellar and panoramic cameras will be 
put on board drones independent of the space 
shuttle. 
As a matter of fact, the major problems to be 
encountered in photographing will be political, 
as certain countries object to their territories 
being photographed without control. 
Mr. Thompson spoke on automated techniques 
in map revision. Though stating that no auto- 
mated systems existed, he discussed digitization 
at the U.S. Geological Survey including digitizing 
old maps, generation of digital profiles, making 
of orthophotos from the digital profiles, and 
digitization of planimetry from the orthophotos. 
He felt that the orthophoto map was the cheapest 
and most rapidly produced revision (or up-date) 
product. 
He also observed that automated techniques in 
map revision were certainly a current object of 
research, but that so far there was little to offer 
in practice, despite an ever increasing demand. 
Here cartographic automation as such is com- 
bined with research for new information and with 
incorporating this information: this is how revision 
differs from up-dating which, again, means a 
complete renewal of the existing map. The 
process of automated revision includes digitizing 
new information (provided by orthophotography, 
for example) and its introduction into existing 
maps: the system called DICOMED (for which 
experiments have been made at USA ETL, Fort- 
Belvoir, U.S.A.) makes it possible to observe and 
to manipulate on the same cathod screen infor- 
mation provided by both an existing map and new 
photography. As the intervention of an operator 
is necessary here, we cannot talk about complete 
automation; what we have is semi-automation. 
Mr. van Zuylen said that, so far, only classical 
methods had been used in the revision of the 
109 
rrr TUE EEE ie