which would cover large areas at a reasonably high speed, with reduction
in the density of geodetic control.
Many systems have been employed and proposed in the last two
decades. Some of them have indeed provided a source of data for mapping
at small scales, but no space imagery available for civilian purposes
has yet proved good enough for mapping at medium scales.
Perhaps none of the existing systems can offer the final solution to
the problem of data acquisition, but the combination of various systems
could lead to an acceptable coverage of the world's demand for maps.
The end of this decade is seeing new technological developments in
this field and most certainly the next decade will see better, faster
and perhaps cheaper methods of acquiring data and producing and/or
revising topographic maps at the scales of 1:50 000 and 1:100 000 so
much in need by many countries.
This paper contains a compilation of the methodology employed and
results achieved by this author and presented in her Ph.D thesis
(Meneguette, 1987). The study aims at employing some of the existing
image forming systems as a source of data for mapping as well as at
developing new techniques for digital map compilation and map revision
at medium and small scales.
Special attention is paid to geometric accuracy obtained by the use
of photographic and scanning systems, but the emphasis of this work is
on the map information content of space imagery and the cartographic
accuracy which can be obtained from line graph and image maps.
2. Methodology
The use of ground control is probably the surest and most accurate
way of orienting images to the ground. In this work coordinates of
selected control points have been read off 1:25 000 and 1:50 000 scale
maps.
A suite of programs developed by this author allowed the conversion
of such coordinates into any desired coordinate system. Additional
control in some test areas was also supplied by mapping agencies (IGN-
France, Ordnance Survey-UK).
In order to keep a digital record of the ground control points used
in this study, files were created, containing the points reference
number, followed by the coordinates given in any chosen system.
In addition to ground control in form of control points, Digital
Terrain Models (DIMs) were generated from aerial photographs and
existing topographic maps were vector digitised.
The resulting database contains most of the information displayed in
the map, namely, main roads, minor roads, railways, rivers, lakes,
coast-line, streets, canals, footpaths, car tracks, reservoirs, salt
pans, etc. All these features have been digitised separately, forming
files with points, strings, links and polygons, containing specific
information.
Vith respect to the digital terrain models, the method adopted in
this study for acquisition of data has been the generation of regular
grids, using newly developed software in a Kern DSR1 photogrammetric
analytical stereoplotter.
This DIM data has been processed by employing subroutines written by
this author, in order to carry out data compression by discarding all
information that can be retrieved at a later stage. Height interpolation
routines have been implemented and sorting of data programs written to
access the data quickly.
247