ProFORMA

[Sreepathi Pai]

For all you 3D modellers out here:

<http://www.youtube.com/watch?v=vEOmzjImsVc>

I'd wanted to build a laser-based system while in College, but that
never came to be. This system uses a simple camera!

--
Sreepathi Pai

[Savio Monteiro]

beautiful ... i thought of using ultrasonic :P but this is definitely better ... visible range ... 

--
Email to group :  dbit-t...@googlegroups.com
To unsubscribe, send email to
dbit-techtal...@googlegroups.com
Visit this group at
http://groups.google.com/group/dbit-techtalk?hl=en

[Savio Monteiro]

don't you think this will make it easy to capture 3d models of humans and their gestures etc  ... no need for motion capture at all ...  get them in the form of iframes and interpolate ... any other suggestions?

[Niranjan Mujumdar]

Hi

>>> For all you 3D modellers out here:
>>>
>>> <http://www.youtube.com/watch?v=vEOmzjImsVc>

That is a very nice system.

This is very similar to a SIGGRAPH paper we found when we were
searching for project ideas in the summer. Its called VideoTrace. What
it does is that it takes videos and builds realistic 3D models of the
objects in the video. I found it very interesting and wanted to
implement (at least a part of it) as part of the final year project,
but then found another equally interesting project which we chose.

I think something on that lines or ProFORMA would be a very good
project to implement, difficult and challenging.

Here's the link for Video Trace :
Paper : http://www.acvt.com.au/research/videotrace/Siggraph07-van_den_hengel.pdf
Proj Website : http://www.acvt.com.au/research/videotrace/

> don't you think this will make it easy to capture 3d models of humans and
> their gestures etc ... no need for motion capture at all ... get them in
> the form of iframes and interpolate ... any other suggestions?

Hmm, i think it would be difficult to do it for humans because of the
complexity of the curves and surfaces involved, and generating a frame
structure for human body might not relate directly to the system
shown. Capturing gestures can be even more difficult, though i
remember seeing another siggraph paper which converted human face
gestures into graphical smileys.

Niranjan

[joy chalissery]

i am currently not into advanced modelling and animation but a beginner with some experience on maya,3dmax,blender,houdini,and zbrush
I hope this might help u all.
http://mi.eng.cam.ac.uk/~qp202/my_papers/BMVC09/BMVC09.pdf

On Fri, Nov 13, 2009 at 12:21 AM, Savio Monteiro <savi...@gmail.com> wrote:

--
Failure is not Defeat but it just means retreating and coming out again with bigger and Better Arsenal -- joy chalissery

[Yorrick Mendes]

This even captures the textures and UV's! its a pretty impressive system but i'm wondering how they dont have more control over the lighting and background while taking the video. i'm guessing they actually captured it some other time without the hand etc and did a recreation for the video. I think the system would proly work for the human body and will only have problems with concave surfaces like eye sockets etc, Eventually you would want to capture with different lightsources also to get the high poly / normal maps and stuff because the camera wont be able to pick small grooves too.

very interesting video tho!

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

--- On Fri, 11/13/09, Savio Monteiro <savi...@gmail.com> wrote: