If the image qualities differ a factor two, eye-strain becomes
noticeable, but can it be overcome by training? The accuracy drops
rapidly if the two images differ more than a factor four in image
quality, as we found in recent research (Hempenius 1968).
The same questions can be asked for illumination intensity differences,
with consequences for the necessary range of adjustment on plotting
instruments, or automatic regulations.
Stereo-fusion between à negative and a positive copy of that |
negative is normally not possible. (See, however, Davson 1962, ps2 14).
This problem is met in practice with side looking radar pictures, they
often differ as "day and night". Can fusion be trained, or is the
application of an alternating illumination and after-image effects a ;
way out? Negative or positive after-images appear after a certain a. s
time lag.
Binocular Vision j = Vision + Stereoscobic Vision
With illumination in complementary colours, the fused model is seen
as white (Kelsh-plotter), which is & peculiar phenomenon. But fusing
a normal colour print with & false-colour print, one might expect the
strong reddish colour to be dominant over the soft green colour, but © ®
many operators see it the other way around. The result is almost the k
same as when viewing a normal colour photograph and a black and white
print of the same area: the model is coloured at first sight, but the
colour fades away after some time. However, some operators manage to
see a coloured model if they want so, but then the height differences
are not noticed as strongly as when the colour appears.
Image contents and the influence of regular structures in the images
was studied by Makarovid (Makarovid 1966).]
Probably a general CORRELATION FUNCTION can be developed for the above
cases. One of the difficulties is felt to be in the limited and variable
amount of image details which is used by the human operators possibly he
restricts to only a very small area for the final setting, with a radius of
2 to 5 times the MM diameter. (De Haas, private comm. 1967).