8. Incorporation of Transfer Function Measurement in
"Recommendations for Camera Calibration"
The Resolution 1.5 envisages "an agreed method" to be
incorporated in ISP "Recommendations for Camera Calibration"
by the 1968 Congress. The Working Group does not feel that
it can recommend any particular method to the exclusion of
others, because from the fundamental nature of the transfer
function the particular method employed is not important,
all methods being equivalent in principle. It is of far
greater importance to ensure that whatever method is employed
should be thoroughly tested and qualified under the widest
possible range of conditions before it is used, and that a
satisfactory proof of accuracy be available. The actual
method employed is then a matter for the choice of each
laboratory.
Since the "Recommendations" is not a standard and
does not include limits or suggested performance standards,
the question of including recommended transfer functions for
particular applications does not arise. It is suggested that
to introduce quantitative notions into general circulation,
typical transfer functions of known excellent lenses, measured
according to the best known methods, should be published, for
example in Photogrammetria.
The Working Group has composed a suggested section, to
be incorporated as a supplement in "Recommendations", dealing
with precautions, details of test techniques, numbers and
types of measurements, and expression of results. This supp
lement is offered tentatively, for incorporation only if
approved after wider circulation and discussion in the meet
ings of Commission I. The suggested supplement will be found
in Appendix 5*
9. Interpretation and Use of Transfer Function Measurements
It must be assumed that in due course OTF measurement
techniques will be developed to an acceptable level of acc
uracy and standardization. The question then arises, as to
how the measurements are to be interpreted and applied for
the purposes of aerial photography.
Application even to the limited objective of specifying
the performance of lenses on an accepted quality scale,
cannot be entirely divorced from interpretation. At present
the user of a lens can obtain a performance figure (resolving
power) which indicates the limiting size of a standard target
that the lens will reproduce in combination with a specified
emulsion, thus providing an intuitively acceptable bridge