Re: [Bf-committers] Code donation: Dual contouring re-meshing

2011-03-22 Thread Ρυακιωτάκης Αντώνης
Wow...just...wow!
___
Bf-committers mailing list
Bf-committers@blender.org
http://lists.blender.org/mailman/listinfo/bf-committers


Re: [Bf-committers] Code donation: Dual contouring re-meshing

2011-03-21 Thread Nicholas Bishop
Quick update,

New video: http://vimeo.com/21330126
Patch against trunk is now uploaded: http://www.pasteall.org/20117/diff

-Nicholas

On Mon, Mar 14, 2011 at 10:19 PM, Nicholas Bishop
nicholasbis...@gmail.com wrote:
 Hi again, some quick status updates on this:

 I've started checking out the code, looks like integration as a
 modifier shouldn't be too hard.

 As a first step, since the code was written for 32-bit compilation
 with MSVC, I've been updating it to compile with g++ too, as well as
 some memory fixes for 64-bit.

 Current status, I've got it working as a standalone program. Quick
 test with the Standford bunny, imported resulting .ply model into
 Blender:
 http://www.pasteall.org/pic/show.php?id=9959
 [on right is original with holes and uneven topology, then middle is
 rebuilt with a depth-7 octree, and on the left rebuilt with a depth-5
 octree.]
 The output looks good to me, and pretty quick to calculate :)

 Right now I'm continuing to update the code externally from Blender,
 making it a bit more cross-platform and reformatting the precalculated
 text tables into human-readable (i.e. not datatoc'd) C files. Once
 this is done I'll pull it into Blender intern/, put together a C
 wrapper, and try it out as a modifier.

 -Nicholas

 On Sat, Mar 12, 2011 at 1:25 AM, Nicholas Bishop
 nicholasbis...@gmail.com wrote:
 Just FYI, I've started looking at this stuff. Nothing to report yet though :)

 Anyone else looking at it?

 -Nicholas

 On Mon, Mar 7, 2011 at 9:19 AM, Sergey Kurdakov sergey.fo...@gmail.com 
 wrote:
 Hi All,

 though I'm not qualified to keep with the donated code, few thoughts,maybe
 they could be of some use

 it looks like
 the approach  will be great to be combined with

 http://www.blendernation.com/2010/11/26/microsoft-kinect-in-blender-realtime-point-cloud-demonstration/

 http://www.blendernation.com/2010/11/26/microsoft-kinect-in-blender-realtime-point-cloud-demonstration/kinect
 like devices get more penetration ( ms will issue sdk and also asus will
 have their device )
 combination of point cloud ( which with more kinects will be very common for
 designers ) with blender
 could be a very good point.

 as for constructive solid geometry there is some code to be used
 http://www.opencsg.org (a link
 was mentioned on list few years back btw )

 and finally  maybe it can be used for fast and accurate booleans
 something like described here:
 http://www2.mae.cuhk.edu.hk/~cwang/pubs/TVCGMeshBoolean.pdf


 Regards
 Sergey
 ___
 Bf-committers mailing list
 Bf-committers@blender.org
 http://lists.blender.org/mailman/listinfo/bf-committers



___
Bf-committers mailing list
Bf-committers@blender.org
http://lists.blender.org/mailman/listinfo/bf-committers


Re: [Bf-committers] Code donation: Dual contouring re-meshing

2011-03-14 Thread Nicholas Bishop
Hi again, some quick status updates on this:

I've started checking out the code, looks like integration as a
modifier shouldn't be too hard.

As a first step, since the code was written for 32-bit compilation
with MSVC, I've been updating it to compile with g++ too, as well as
some memory fixes for 64-bit.

Current status, I've got it working as a standalone program. Quick
test with the Standford bunny, imported resulting .ply model into
Blender:
http://www.pasteall.org/pic/show.php?id=9959
[on right is original with holes and uneven topology, then middle is
rebuilt with a depth-7 octree, and on the left rebuilt with a depth-5
octree.]
The output looks good to me, and pretty quick to calculate :)

Right now I'm continuing to update the code externally from Blender,
making it a bit more cross-platform and reformatting the precalculated
text tables into human-readable (i.e. not datatoc'd) C files. Once
this is done I'll pull it into Blender intern/, put together a C
wrapper, and try it out as a modifier.

-Nicholas

On Sat, Mar 12, 2011 at 1:25 AM, Nicholas Bishop
nicholasbis...@gmail.com wrote:
 Just FYI, I've started looking at this stuff. Nothing to report yet though :)

 Anyone else looking at it?

 -Nicholas

 On Mon, Mar 7, 2011 at 9:19 AM, Sergey Kurdakov sergey.fo...@gmail.com 
 wrote:
 Hi All,

 though I'm not qualified to keep with the donated code, few thoughts,maybe
 they could be of some use

 it looks like
 the approach  will be great to be combined with

 http://www.blendernation.com/2010/11/26/microsoft-kinect-in-blender-realtime-point-cloud-demonstration/

 http://www.blendernation.com/2010/11/26/microsoft-kinect-in-blender-realtime-point-cloud-demonstration/kinect
 like devices get more penetration ( ms will issue sdk and also asus will
 have their device )
 combination of point cloud ( which with more kinects will be very common for
 designers ) with blender
 could be a very good point.

 as for constructive solid geometry there is some code to be used
 http://www.opencsg.org (a link
 was mentioned on list few years back btw )

 and finally  maybe it can be used for fast and accurate booleans
 something like described here:
 http://www2.mae.cuhk.edu.hk/~cwang/pubs/TVCGMeshBoolean.pdf


 Regards
 Sergey
 ___
 Bf-committers mailing list
 Bf-committers@blender.org
 http://lists.blender.org/mailman/listinfo/bf-committers


___
Bf-committers mailing list
Bf-committers@blender.org
http://lists.blender.org/mailman/listinfo/bf-committers


[Bf-committers] Code donation: Dual contouring re-meshing

2011-03-07 Thread Ton Roosendaal
Hi all,

I've been contacted by Prof. Tao Ju, University St Louis.
http://www1.cse.wustl.edu/~taoju/

He offered Blender Foundation to use algorithms and code he developed  
the past years. It's based on his 2002 Siggraph paper:
http://www1.cse.wustl.edu/~taoju/research/dualContour.pdf

He further worked on that, resulting in this program:
http://www1.cse.wustl.edu/~taoju/code/polymender.htm

It also has an improved method for creating meshes out of volumes,  
preserving detail:

example of a marching cube metaball scene:
http://reocities.com/SiliconValley/port/3456/isosurf/png/demo1.png

the same metaball scene in dual contouring:
http://reocities.com/SiliconValley/port/3456/isosurf/png/demo2.png

His assistant Jean-Denis Boudreault wrote me:

If blender wishes to push the envelope even further, it could enter a  
whole new area of modeling paradigm, in which dual contouring would be  
the key. for example, the 3d modeling software http://www.3d-coat.com/  
uses a voxel data approach for advanced neo-modeling of polygon free  
meshes. Blender could enhance it's metaball system with dual  
contouring to make it's entry in the voxel based modeling world in  
style.  all this is only the start an overhaul of the metaball system  
can offer. With dual contouring, you now have the proper tools to do  
it.

When I wrote back about possible implementation of CSG meshing, he  
wrote:

As for CSG operations, I was not aware that they were not included in  
the above code. But im sure the talented Blender developers shall know  
how to re implement it if necessary. If this can help, i include code  
from an independent experimental implementation of dual contouring  
that is available as open source on the internet 
(http://reocities.com/SiliconValley/port/3456/isosurf/isosurfaces.html 
). Now i am not aware if the code is usable by blender as is (and im  
convinced Mr. Ju's code is preferable anyways), but it can surely be  
used as an example implementation of CSG operations, from which the  
developers can triangulate the algorithms into Mr. Ju's code.

The code for dualcontouring and polymender is here:
http://download.blender.org/demo/test/dc.zip

It is still (c) by the creator. He has agreed on transferring to this  
to (c) for BF, but with keeping his contribution credit.

Anyone interested in following up on this is welcome! I'd need  
feedback from experienced developers what do with it precisely. In  
case there's someone who can (wants to) bring this in Blender, Tao Ju  
is more than willing to assist or give feedback too.

Thanks,

-Ton-


Ton Roosendaal  Blender Foundation   t...@blender.orgwww.blender.org
Blender Institute   Entrepotdok 57A  1018AD Amsterdam   The Netherlands

___
Bf-committers mailing list
Bf-committers@blender.org
http://lists.blender.org/mailman/listinfo/bf-committers