les
the
it is
of
angle
rough
der-
area
is
es
eny
t be
ints.
tri-
do-
f the
en,
nbi-
the
the
RE
[Cy
jer-
EL y
3 im-
LC
ak.
^"0-
for
.able
ten-
ind-
lese
igh
l-
ly
ure
irst
ete
of
ul
pre-
nal
ur-
pro-
1g=
rder
18
> be
cs,
on-
di-
^k
as for example mountain ridges and axis of cur-
rents for the case of hydrographic net, informa-
tion for road construction etc. These can be ta-
ken from the alignment which is almost always de-
sirable at photogrammetric practice or they may be
drawn especially for this reason.
4. CALCULATION OF SECTIONS AND CONTOUR CURVES
The calculation of the sections is based on the
calculation of the altitudes of lonely points. We
can be interested in a section along specific po-
ints or along a straight line every determined
step. At both the cases, from an archive of form
code X, Y we end in an archive code S, Z where S
is the distance from the beginning along the sec-
tion and Z is the altitude.
The calculation of contours concerns in the begin-
ning the search of triangles of the polyhedral
surface which are cut by the horizontal level, the
correspondent to the altitude of the contour. Du-
ring the calculation procedure in the beginning,
we search for every triangle if it is cut by the
horizontal level which corresponds to every con-
tour, and in this case we search on its sides for
points of altitude corresponding to the one of the
contour and we determine the straight line that
comes from these points. The archive of the strai-
ght lines that appears is not classified because
the triangles are taken from the archive of the
surface in series. Suitable algorithm, which allo-
ws the classification of the straight lines, has
been developed. This allows among others the con-
tineous drawing of the contour. At the same time,
a smoothing of the contour, which has in the be-
ginning a polygonal form because of the way of its
formation, is achieved. The smoothing can be at
first achieved by several techniques. In this pa-
per polynomial of second grade was used for smooth-
ing.
5. EXAMPLE
All that was mentioned in this paper were applied
at the study of a drainage net. The study was done
based on airphotographs of scale 1:30,000 (The
photogrammetric works were done with Topocart
Iena).
Contours (Fig. 3) and sections (Fig. 4,5) were the
products. A usual photogrammetric restitution of
the area (details of alignment, contours) was con-
Sidered intentional (Fig. 1).
6. CONCLUSIONS
The subjective factor is very important at all the
procedure of the choice of sampling points. This,
of course, is followed by the weakness which is
presented by this consideration in connection with
sampling on some density or other procedures.
However, aspects as the mentioned ones relatively
to the sampling procedure, cover at great extent
the weakness that inevitably exist because of the
nature of the sampling procedure.
Several methods are developed to control the accu-
racy of DTMs. Sampling methods and geomorphologi-
cal considerations are basic factors which deter-
mine the accuracy. Besides, the scale and the use
of DTM are realistic considerations, which deter-
mine the grade at which the DTM may be acceptable.
In this paper which is a first presentation of the
DTM, a systematic control of the accuracy was not
done. A qualitative evaluation of it may be consi-
825
dered to come from the comparative consideration
of the form of the contours, resulting from DTM
and from the photogrammetric restitution. This
comparison may be satisfactory in the beginning.
The application of the presented DTM was done on
base of airphotographs of relatively small scale
of a.small area in order to have a first idea of
its possibilities and its restrictions. However,
its further use, for example from a Service (vari-
ous instruments-operators, large number of models,
several applications) could offer valuable conclu-
sions for its practical use.
The structure of the developed DTM makes it suita-
ble for general use (different landforms and appli-
cations). However, a particular facing, mainly at
the stage of sampling, which will be indicated by
the kind of the application (e.g geomorphological-
hydraulic subjects, road construction etc.), could
be specially useful for the quality of all the
work.
Especially concerning the study of the drainage net
we have to remark that it is particularly important
as landform because all that happens in a drainage
basin (erosion, landslide, floods etc.) decisively-
multiply influence the whole area (inhabited areas,
coasts etc.). In Greece, because of its characteri-
stic relief, a multiple study of the drainage net
is particularly useful and necessary in order to
have a successive and complete geo-information sy-
stem and further more for the successive applica-
tion and exploitation of every developing program.
The study of the parameters that are being used for
geomorphological and hudraulic studies can be sa-
tisfied at great extent by the presented digital
model.
We notice the possibility of taking a variety of
other products beyond contours and section (per-
spectives, volumes etc.) which may be taken on the
base of the above digital models with suitable
additions to the program, according to the research
that might interest. Every effort was made in order
to have the whole work to the direction of timely
aspects about DTM ( WG III/3, DTM, ISPRS August
1986, Rovaniemi, Finland), relatively to the resea-
rch of methods which better describe the forms and
they present possibilities of practices and may
cover large areas (Data bank on national level).
REFERENCES
1. Makarovic B., 1973, "Progressive sampling for
digital terrain models" I.T.C Journal 1973-3,
The Netherlands.
2. Makarovic B., 1979, "From Progressive to Compo-
site Sampling for Digital Terrain Models" Geo-
Processing 1 (1979), pp 145-166, The Netherlands.
3. Patmios E., Tsakiri-Strati M., 1982, "DEM's
Studies For Information About Types of Terrain"
ISPRS Com. III, Mathematical Models, Accuracv
Aspects and Quality control, June 7-11; Helsin-
ki, Finland.
4. Paresi C., 1982, "Modéles numériques de terrain"
I.T.C The Netherlands.
