mpling points, we can calculate its altitude Z, as
section of the vertical line which crosses it,
with the corresponding triangle of the polyhedral
surface of the model.
If Xi, Yi» 7i, Xj» Yj, Z5, Xy» Yy, Z, are the co-
ordinates of the apexes of a ijk triangle of the
polyhedral surface of the model, the equations
AXj + BYj + CZi- 1
AX; + BY; + CZ j- 1
AX, + BY, + CZy- 1
describe the level of this triangle (A, B, C are
the coefficients of its direction).
Hog
99909
The altitude Zp of a point with horizontal coordi-
nates Xp, Yp which is in the projection of the
triangle ijk, is given according to the above,
from
Zp ed AXp - BYp MG
The sampling points are chosen by a suitable pro-
cedure for the most possible true presentation of
the surface of the land.
From the sampling points Pi, of which the coordi-
nates Xpi, Ypi, Zpj are the input data of the mo-
del, suitable triangles at space are formed so
that they describe most truly the surface of the
land.
This need leads to a procedure of Seeking the most
suitable connection every two projections Xpi, Ypi
of the sampling points Pj at the horizontal level
XY in order to form at the level XY the most suit-
able triangles of which the corresponding ones at
the space, satisfy the relevant demands.
This searching through in some Way corresponds to
the procedure in topographic planning for the
drawing of the contour curves.
The connection among the points must be suitably
done so that to form a homogeneous net of trian-
gles.
The choice of the connections is of decisive im-
portance for the general quality of the modes and
it is closely connected with the sampling proce-
dure. However, connections among points which are
generally far away from each other, must not be
taken, because, except others, they could omit
details of the forms. Besides, the straight lines
of the connections must not cut each other. Gene-
rally the length of the straight lines of the con-
nections must be short so that to connect neigh-
bouring points among which the considerations of
linearity are closely to the reality.
Aspects as the above and others relevant to the
volume of the calculating procedure, to the possi-
bilities of memory of the computer etc. led us to
search through the most suitable connections in
smaller areas. These areas were thought useful to
be considered as circular surfaces centered at
each point. The radius of these surfaces depend on
the in general distribution of sapling points.
The connections of every point-center with the
included points in the corresponding circular sur-
face are searched through, on the base of the abo-
ve demands.
This searching through finally gives a totality
of acceptable connections among the sampling po-
ints from which the triangles of the polyhedral
surface of the model are formed with suitable
procedure.
824
The coefficients of the level of these triangles
are calculated and recorded at the archive of the
Surface. After the creation of this archive, it is
possible to find, as mentioned, the altitude of
any P (Xp, Yp) point. For this reason the triangle
at which the point corresponds is searched through
in the beginning. This is easier on the consider-
ation of a grid at XY level which covers the area
of the surface of the model and therefore it is
possible for this to be done only for triangles
which are completely or partly included at every
cell on the grid. The density of the grid must be
in connection with the density of sampling points.
Generally, a thick grid limits the number of tri-
angles in which the searching through will be do-
ne, but it, of course, increases the number of the
cells.
The way in which the archives have been written,
the series of the relevant works in such a way
that the presuppositions, that were put in combi-
nation with time saving, to be secured, were the
objects of suitable calculating procedures of the
constructed program.
3. BASIC PRINCIPLES OF THE SAMPLING PROCEDURE
The choice of sampling points at some regularity
essentially serves automatism, releases the oper-
ator from the care of their choice and generally
standardizes in a way the work. However, it is im-
possible to prevent from loosing characteristic
points, influencing the quality of all the work.
Methods of progressive sampling or compound pro-
cedures as mentioned, are particularly useful for
geomorphological aspects and for offering reliable
elements for applications. However, Special atten-
tion and high calculating substructure is demand-
ed at the various parts of the procedure of these
methods.
In relation to the above, the presented in this
paper sampling procedure is characterised by high
demands of succesive subjective choice of sampl-
ing points so that the form of the land is truly
described.
From some aspects it corresponds to the procedure
followed by the classic topographic works.
The stereoscopic observation of the model at first
with mirror stereoscope and the possible complete
conception of the relief determine the quality of
all the study. This study is particularly useful
to be followed by simple drawing of photointerpre-
tation details at the airphotograph, the original
or magnification, which the operator may use dur-
ing the taking of sampling points at the instru-
ment.
The scale of airphotographs, the scale of the pro-
duced map and the density of points that are sug-
gested by the scale of map and the category of
terrain, must be considered in connection in order
to form an opinion about the density of sampling
points.
Characteristic alignment elements are useful to be
thought of, during the choice of sampling points,
because the taking of sampling points may be con-
sidered intentional along them. These alignment
elements may also help the choice and the defini-
tion of smaller areas in which the sampling work
is indicated to be gradually done. They may vary
Lh FX go FRÜDO O e nÉn SS ID HAMM OOS "Ys: OU r1 tHE OME OO 13
124.4 "wv*4 7 n3 "m
FA i rot Nh
