RELATIVE ORIENTATION IN MOUNTAINOUS TERRAIN, DISCUSSION
Author’s Presentation of the Paper
Mr A. J. VAN DER WEELE: After preparing
a first draft of my paper about Relative Orien-
tation in Mountainous Terrain I got reactions
from several people, including Professor Bach-
mann, Mr Hallert and Mr Wassef. I thank them
very much for their contributions.
As an introduction I should like to start by
pointing out that the method of relative orien-
tation consists in the first place of eliminating
y-parallaxes in the model, but there is also the
possibility of trying to improve the result by
including elevations or x-parallaxes. In my
paper, in the first place, I have given a review
of existing methods for the elimination of y-
parallaxes. This paper has no pretention of
being original or of giving anything new; it
simply tries to indicate the methods which have
been published already on former occasions.
To simplify matters, I have tried to show
that the parallax formula for all types of photo-
graphs can be written in the same form con-
taining five terms which are valid for every
type of photography. If those five terms are
meant to determine elements, these may not be
elements which you can use with instruments
but they could be called shadow elements or
geometrical quantities (parameters).
In the second place, I have indicated that
the difference between mountainous or flat ter-
rain lies mostly in the determination of ww.
There are a number of graphical and numerical
methods for the determination of Aw in moun-
tainous terrain, and I have mentioned some of
them in my paper.
I do not think there is much to say about
relative orientation as far as the use of y-paral-
laxes is concerned. The more difficult point is
the inclusion of x-parallaxes for an improve-
ment of the orientation. The equation which
governs this procedure contains seven unknowns,
so that, in principle, seven parallaxes must be
known to solve the problem completely. One
of the unknowns being the scale factor, we can
eliminate that by using situation points, so that
six unknowns remain. Those six unknowns
include elements of relative orientation as well
as elements of absolute orientation.
In considering this problem I have had the
impression that a general solution, as is pos-
sible for the elimination of y-parallaxes, can-
not be found in the case of using x-parallaxes,
simply because of the fact that we are not free
in the choice of the position of those points.
The position depends on the fact where eleva-
tion points have been determined on the terrain
and the corresponding positions of those points
on the image. Therefore, it will be necessary
for every special case to solve the six equations
if you put the problem in the mathematical
sense, or if you apply empirical methods of
orientation if you want to do it in the instru-
ment.
The procedure to be used depends com-
pletely on the position and elevation of those
points, so that my conclusion is that a general
solution does not exist. Besides, there is one
point which I neglected in my paper: that is,
the influence of elements of interior which, in
the case of mountainous terrain, will have a
great influence on the parallaxes and model
deformations. Including those unknowns in the
problem will make it more complicated. This,
to my view, is a support for my conclusion that
so far a general solution is impossible.
During this Congress, however, a few other
thoughts have occurred to me about this prob-
lem, especially with respect to the type of in-
strument which has been discussed by Mr
Helava in this case, which includes, in fact, a
computer in a restitution instrument. In that
case, there will, indeed, be a possibility of
including a programme in the computer which
easily solves the number of equations which is
necessary and which might give an improve-
ment of the absolute orientation as far as the
elevations are concerned. If there is a pos-
sibility of disposing of an electronic computer
separate from a normal instrument, in that case
also a practical solution might be possible,
although at the moment I do not consider this
a very economical solution.
The content of my paper is rather technical.
It includes a great many formulae; I will not
bother you with demonstrating them.
[ think after this introduction you will be
sufficiently aware of the problems with which
we are faced.
Discussion
Mr A. J. VAN DER WEELE: The first thing I
should like to do is to ask if there is someone
on the floor who might be inclined to make
some comments on the paper which has just
been introduced to you. As this seems not to be
the case, I should like to call on Mr Wassef
to £
bee
tha
wh
Thi
fun
at t
ang
thes
The
the
invi
the
rap
give
eith
sec
exp
fore
par
regi
enc
ang
tior
exp
geo
dire
syst
the
sm:
the:
red
to 1
the
onl
in t
dire
em
pro
Ori
mo
ysis
eac
exp
anc
ina
tag
con
exp
ori
tha