englim (englim++at++englim.singapore.sgi.com)
Mon, 31 May 1999 07:04:12 +0800
> Michael T. Jones wrote ...
> (... trimmed ...)
>
>It seems that you are not satisfied with converting the points into
>a pfGeoSet and drawing that, but if you did, and you move the
>resulting point-surface interactively, you will be able to see and
>understand the data. The human mind is quite amazing when
>interpreting moving correlated images...
Yes, with such a powerful feature in our Visual Perception system,
I had wondered whether there were studies made about it. And
as such, whether attempts were made at giving this feature a name.
The closest I came to, was the name 3D-SFM (3D-Structures
From Motion). In "The Vision of the Brain" by Semir Zeki, a simple
but compelling experiment describing this feature was included.
What initially seemed to be a 2D-disc, full of red dots over a green
background, is suddenly perceived as a 3D-globe, when the dots
were moved coherently in a certain way (rotated).
More compelling when demo'ed, was that the 3D globe (structure)
is perceivable only if the red & green colors are of different
luminances. When they were made to be of equal luminance, the
globe is no longer perceivable.
Something to note when building effective real-time visual systems?
Cheers,
Eng Lim Goh
SGI
This archive was generated by hypermail 2.0b2 on Sun May 30 1999 - 15:59:21 PDT