[photo-3d] RE: possible discussion topic: stereocorrelation

["Sergio Baldissara" <winter@xxxxxxxxxx>]

Hi Greg! (& everybody)
Indeed I believe stereocorrelation SHOULD  be the RIGHT way to cast so-said
3d cybermodels.
The state-of-art for 3d modelers are softwares like CANOMA or the free
POV-Ray (they allow the user to add arbitrary depth) or  tools like
http://www.minolta.com/dp/3d1500/ , whose software re-builds 3d (nonstereo)
models in the way I (badly) described in
http://www.pauck.de/archive/mailinglist/photo-3d/mhonarc/msg38649.html .
I believe 3d modelers choose the "wrong way" because:
1. none of them came from 3d photography;
2. human brain's computing speed in 3d modeling from  stereo vision is still
too hard to approach.
Sergio

----- Original Message ----- > >
>    From: Gregory Camilli <gac@xxxxxxxxx>
> Subject: possible discussion topic: stereocorrelation
> I've only joined recently so this request may not be appropriate for this
group. I've
> recently come across the topic of stereo correlation and have located a
product
> (camera + software) that uses a stereocorrelation algorithm to create 3-D
point
> clouds.  This seems like an intriguing methods for creating 3-D models
from
> stereoscopic photographs.  However, I've had no luck in locating readable
material on
> how the algorithms work.  I'm wondering if anyone here can shed some light
onm the
> subject.
>