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]

T3D Re: TECH-3D digest 362 / mounting quality

  • From: Rolf Henkel <henkel@xxxxxxxxxxxxxxxxxxxx>
  • Subject: T3D Re: TECH-3D digest 362 / mounting quality
  • Date: Tue, 22 Sep 1998 11:19:50 +0200

--------------492E31ED7DF69F48B8240DBD
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Just a short note to Tony Alderson's:

> Has anyone ever done surveys of the quality of mounting in stereo
> competitions?  I'm not talking about what we "ought" to do, or what our
> goals are, but what is the actual practice.  How much deviation is really
> on the slides?

I'm using old & new stereo photographs for testing of stereo algorithms and was
amazed to see some large alignment deviations in these stereophotographs.
Vertical errors of about 2% of the image sizes, and tilt errors of about 0.5
degrees seem to be quite common. These values were found by an automatic
vergence system.

It seems that humans normally tolerate all these deviations - even small scale
changes between the images of a stereo pair are not directly visible. I guess
the human vergence system is quite effective in canceling all these deviations.

Rolf Henkel

--
 Institute of Theoretical Neurophysics, University Bremen, Germany
 - Email:       henkel@xxxxxxxxxxxxxxxxxxxx
 - URL:         http://axon.physik.uni-bremen.de/



--------------492E31ED7DF69F48B8240DBD
Content-Type: text/html; charset=us-ascii
Content-Transfer-Encoding: 7bit

<HTML>
Just a short note to Tony Alderson's:
<BLOCKQUOTE TYPE=CITE>Has anyone ever done surveys of the quality of mounting
in stereo
<BR>competitions?&nbsp; I'm not talking about what we "ought" to do, or
what our
<BR>goals are, but what is the actual practice.&nbsp; How much deviation
is really
<BR>on the slides?</BLOCKQUOTE>
I'm using old &amp;&nbsp;new stereo photographs for testing of stereo algorithms
and was amazed to see some large alignment deviations in these stereophotographs.
Vertical errors of about 2% of the image sizes, and tilt errors of about
0.5 degrees seem to be quite common. These values were found by an automatic
vergence system.

<P>It seems that humans normally tolerate all these deviations - even small
scale changes between the images of a stereo pair are not directly visible.
I&nbsp;guess the human vergence system is quite effective in canceling
all these deviations.

<P>Rolf Henkel
<PRE>--&nbsp;
&nbsp;Institute of Theoretical Neurophysics, University Bremen, Germany
&nbsp;- Email:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; henkel@xxxxxxxxxxxxxxxxxxxx
&nbsp;- URL:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; <A HREF="http://axon.physik.uni-bremen.de/">http://axon.physik.uni-bremen.de/</A></PRE>
&nbsp;</HTML>

--------------492E31ED7DF69F48B8240DBD--


------------------------------