Mailinglist Archives:
Infrared
Panorama
Photo-3D
Tech-3D
Sell-3D
MF3D

Notice
This mailinglist archive is frozen since May 2001, i.e. it will stay online but will not be updated.
<-- Date Index --> <-- Thread Index --> [Author Index]

[photo-3d] 3D vision


  • From: Pgottfried@xxxxxxx
  • Subject: [photo-3d] 3D vision
  • Date: Wed, 11 Oct 2000 09:48:40 EDT

The ability to see 3D from a stereo pair is only part of the 3D mechanism of 
the brain.  There are four components of depth perception.  The one most 
talked about here is binocular vision.  The brain combines the views from 
each eye into one image after analyzing the differences and determines depth. 
 The second & third component is accommidation.  This means that the brain 
has information of the direction each eye is pointing (i.e. 2nd component) 
and the brain also has information as to the shape of each eyes lens and the 
size of the pupil (i.e. 3rd).  The fourth component is visual perception or 
depth cues.  It is depth cues that allow us to guess at the depth we see in 
2d images and involves things like distant objects becoming grayed or close 
objects appearing bigger than farther objects.  The position of the head, 
especially over time, can also be a factor in depth perception, hence the so 
called "wobble" effect.

When a person has correct vision all the factors are used to see depth.  If a 
person only has one eye or one is weaker then the other, the use of binocular 
is diminished and the others become more dominant.  As stated depth 
perception is a learned response.  the brain correlates all the information 
it has so one perceives the depth.  If binocular vision was the only way to 
perceive depth a one-eyed person could not drive a car or hit a baseball.

In the case of stereo imaging the brain mainly uses binocular vision to see 
depth and must ignore most of the accommidation information, hence why some 
say it gives them a headache.  A one eyed person can still see depth, just 
not the same way.  A one eyed person uses depth cues, accommidation and the 
position of the head over time to perceive depth.

Most of my stereo imaging involves lenticular imaging.  I routinely create 
perspective sequences of 12 to 30 frames and before interlacing I preview 
them as stereo pairs or anaglyphs.  The wider the seperation between frames 
the more I have to ignore accommidation.  

That said, my point is that not all stereo/depth perception is learned at a 
young age.  Optical illusions such as wall-eyed stereo, cross eyed stereo and 
stereograms  are often difficult for some people to see until they have 
trained themselves to ignore certain aspects of the depth peception they 
have.  Those with one weak eye may find stereo images virtually impossible to 
see but still be able to see real world depth.