Game Development Reference
i.e. frequency channels (Campbell & Robson, 1968). The contrast sensitivity function
thus would not reflect the sensitivity of one type of cells, but would be a combination of
the sensitivities of all these filters. A number of aspects of this hypothesis are extremely
interesting and motivate a large number of researchers working on this subject:
Similarity with the auditory system in addition to the idea that the nervous system
probably treats similar problems by using similar methods;
Compression of information: in fact, visual information has a large number of
periodicity (at least locally) which needs to be coded economically since the optic
nerve has only a million fibres for hundred and twenty-six million receptors.
In their article which initially put forward the possibility of filtering of visual informa-
tion depending on its frequency content, Campbell and Robson (1968) measured the
detection of various complex patterns. They linked the detectability of these patterns
to their spectrum amplitude and to the contrast sensitivity function of an observer for
the sinusoidal networks.
To conclude, the reader may find that the detailed study of information process-
ing mechanisms used by the visual system is unrelated to the main concerns of virtual
reality, but we will see that this is not the case. What purpose would be served in cre-
ating an immersion using three-dimensional images if the user is going to have trouble
looking at them after a few seconds? We will illustrate this in the chapter “Stereo-
scopic restitution of vision'' with an application that processes images by wavelets to
improve the perception of three-dimensional images. Wavelet analysis is equally useful
for the perception of depth by the human visual system. But before that we are going
to explain the ways in which a human being perceives depth because a lot of people
still confuse visual perception of depth and three-dimensional or stereoscopic vision.
3.2.2 Visual perception of depth
220.127.116.11 Cognitive perception by monocular cues
It is necessary to first understand that the perception of depth of a three-dimensional
world is not only because of binocular vision. With only one eye, a human being
can interpret the image received and deduce the notions of depth. This interpretation
is done subconsciously through a cognitive processing thanks to what he has learnt
from a very early age. At the early stage of visual processing, the decomposition and
processing of visual information by different channels makes it possible to use a varied
series of cues for the perception of three-dimensional space, and thus that of depth.
These cues can be divided into two main categories:
Proprioceptive cues (caused by actions of orbital and ciliary muscles), made of
accommodation and convergence : these cues are weak at a short distance (a few
Visual cues, made of binocular and monocular cues.
Proprioceptive cues, accommodation and convergence, are adjusted by the visual sys-
tem. It is through the value of command of the orbital and ciliary muscles that the brain
has a proprioceptive perception of the depth of the object observed. Let's not forget