WORKING GROUP 1
WELANDER
29
tial advantages for studying the different links of the image-forming system
where the aerial photograph constitutes the final product. These new aids
- the transfer functions (contrast transfer functions) - were introduced to the
photogrammetrists two years ago at the I.S.P. London congress. There, a few
authors gave reports on the theories, supplemented by some applications
in photogrammetry [1—5].
The Steering Committee of Commission YU I.S.P. has invited the author
to give a general survey of the methods and investigations for determining the
quality of aerial photographs - with emphasis on their practical applications.
For this purpose a comparison between the methods used up to now and the
new transfer functions will be made.
The resolving power
In most parts of the world the well known concept “resolving” power (RP)
or, better, the “resolving limit”, is still used as a quality criterion of an image,
which is either projected or consists of a photograph. However, it was found
quite early, that the values of the RP obtained in evaluating aerial photographs
did not correspond to judgements of the same photos, carried out by skilled
photo interpreters. One of the causes for this was that up to then, high test
contrasts (density differences ~1.3) were used in the RP-determination,
while the contrasts in an aerial photograph are commonly small. Then the
use of low contrast patterns of various appearances was started. These did indeed
give a little better result, but the agreement with the result obtained in practice
was still unsatisfactory.
Commonly, the highest RP was received for thin negatives developed in
comparatively short times. The granularity here is low, but at the same time
the contrasts will be small too. This causes the interpreter to suggest that such
photographs are not the best ones. Several investigations have shown that the
highest RP is obtained at density values of about 0.8-1.0. At higher density
values the grains will be too large.
The acutance
In the meantime a new method had been evolved where the photographic
image of a knife-edge (high contrast) was more closely analysed by scanning
with a micro-photometer [6]. On the basis of this, acutance of the photograph
was calculated in a special manner (fig. 1). In certain cases this value gave a
better agreement than RP, compared with the result for judgement in
practice.
It has to be remembered that for certain classes of lenses the quality of a
photograph varies according to whether the camera is focused for the best RP or
the best sharpness. In the latter case the acutance gives a fairly good agreement
with the visual judgement of a photograph. Attempts were also made to deter
mine the quality of an image as a function of both RP and acutance. None
