nd some Modes
Record, Vol. II,
]g Possible Local
io. 4, September
in Topographic
63.
f Estar Polyester
ng, Vol. XXVII,
STEREOSCOPY AND PHOTOGRAMMETRY
By D. A. Palmer
ional Archives of
National Physical Laboratory
inadiart Surveyor,
(Invited Paper, Commission I, Xth International Congress of Photogrammetry,
s of Atmospheric
XXVI, No. 5,
Lisbon, 1964)
and Hypersonic
Mo. 3, June 1961.
1 and Terrestrial
> the IXth Inter-
Abstract
Recent developments in stereoscopy are reviewed, and their relevance
to photogrammetry discussed.
'es images
s résultats
-ogramme
■■’.SW?:'
Photogrammetry, like many other highly developed scientific techniques, still
relies on human beings for an essential part of its operation, namely those whose
stereoscopic judgement forms the vital link between the photographs and the contour
maps. There have been several attempts to replace the slightly erratic humans by
repeatable and predictable machines, but as yet the consequences have been either
more expensive or less convenient. [1_3]
Meanwhile, studies of binocular vision are also progressing, and the
photogrammetrist should keep one eye on such developments to see if they might
save him time and trouble. Although stereoscopy is a comparatively slow moving
subject, this is not because everything is already known. On the contrary, the
process whereby we are able with two eyes to see objects in three dimensions is still
somewhat obscure. Indeed, at the moment we are able to study only the conse
quences of that ability rather than the ability itself, because so many of the clues lie
in that inaccessible organ, the brain, of which the retina is histologically an extension.
The sine qua non of stereoscopy is that we have two eyes capable of looking
simultaneously in the same direction. Thus an object under observation will form
two images, one on each retina. In the normal observer, the impression is never
theless of a single object; the retinal images have been fused as if according to the
interesting equation
A + A = A
where A represents one image. Incidentally, this curious algebra is valid even
photometrically, as anyone can see for himself, since a scene viewed with two eyes
appears no brighter than when viewed with only one.
In fact, the normal observer (by no means everybody) does obtain something
extra on the right-hand side of the equation, namely the sensation of depth. The
objects may appear to be single, but they are also seen standing out at various
distances within three-dimensional space, a striking experience peculiar to those
with stereoscopic vision.
These simple observations suggest the following questions:
By how much must objects be separated in depth before they become noticeably
so?
391