'Ose
net-
Ish-
na-
, t0
lere
der
s of
the
ira,
d it
ble.
uld
han
this
ible
be
the
Ssi-
Commission IV Invited paper 113
Convergent versus Vertical Photography
by W. SCHERMERHORN
LT.C., Delft.
This title may convey the impression firstly that there is some controversy on this
matter, which is the case, but secondly that it may be intended to pass judgment on which
of the two alternatives is superior to the other, which would be misleading. As with any
other technical aspect of photogrammetry, the only generalisation which is always ap-
plicable is that NO generalisation can be applied to all circumstances. For any compar-
ison of methods to be of value, that comparison must be related to the suitability of the
methods for particular requirements. This will be our approach to the matter.
Consistent with this approach, Prof. Kasper’s suggestion that the question should
rather be: “In which cases can convergent photography have advantages?" has much
merit. I would, however, suggest the addition of the further question: *What are these
advantages?".
A further preliminary remark is that we must be careful to distinguish between
purely geometrical considerations and those which derived from the results of productive
work — precision is important, but so, also, is production time. Base-height ratios and the
area of coverage per model must obviously be studied, but they will not give us the whole
answer to the problem. The relation between base-height ratio and the attainable preci-
sion of height measurement, particularly towards the upper limits of the range of this
ratio, between 1.0 and 1.4, is something on which there is as yet insufficient evidence.
This relation can only be determined empirically from tests. Unfortunately we do not
know such results which can give a reliable answer to these questions. Also the reports
received from the panel members are, from the point of view of this kind of basic obser-
vations and results, rather poor. As long as no better and no more material is available,
decisive conclusions will remain impossible.
While several other criteria may be proposed, the criterion of precision may be con-
sidered from numerous viewpoints which are related to economy:
1. The precision of planimetry, in one model, is generally a decisive factor for large-
scale maps such as those for cadastral purposes.
2. The precision of height measurement and of contouring, in one model, is important
for topographic maps of medium scale and for large-scale mapping for engineering
purposes in terrain which is not flat.
3. The precision attainable in the aerial triangulation of a strip of given length (or a
given number of models) will have a significance, in all types of work where trian-
gulation is used, quite different from the other factors mentioned.
From each of these considerations we can derive a minimum value for the scale of
photography. The largest of these values will determine the necessary scale and thus
have a decisive effect on the economy of the whole procedure.
It will be obvious, however, that a general conclusion cannot be reached by consid-
ering this approach alone, for there will be cases in which criteria which are, more or
less, unconnected with precision, will play a decisive role, as we will see. Even when
some aspects of precision are important, it may be that others do not enter the picture
at all. If we consider, for instance, the photogrammetric production of maps for highway
engineering in entirely flat terrain, for which purpose convergent photography has been
in regular use in the Netherlands since 1932, the construction phase of the project
requires the availability of a network of points known both in planimetry and height,
and there is, thus, no need at all for aerial triangulation. Efficiency in the project is