[x3d-public] simulating surgery > realtime plastics > cuttable, deformable

doug sanden highaspirations at hotmail.com
Sat Feb 4 09:04:17 PST 2017


Looks like X3D is missing softbody dynamics.
https://en.wikipedia.org/wiki/Soft-body_dynamics
...
- can be kind of like our particle physics -you can compose forces, solve with simple equations
- except with springs between points, and stiffness constraints between 3 points in a line, in a grid
http://cg.skeelogy.com/?download=SoftBodyPhysicsTutorial
if you are starting with an empty hull, you would/could 'tetrahedralize' the interior to create interior 'cells'
if you are starting with 3D grid data -3D texture, volume rendering- you can use that more directly
- using something like FEM finite element method http://run.usc.edu/vega/
- perhaps with opencl or shader program for gpu acceleration

In theory irregular surface meshes could be vertex-linked to a regular 3D grid (ie each vertex has a cell index), finite element applied to the grid, and vertices updated (instead of tetrahedralizing the surface mesh directly), and this approach might work best with medical simulations which might have a 2D surface texture/skin, and an 3D volume texture ie MRI/CT scan, segmented on tissue properties ie bone vs .. and different springyness / stiffness constants applied to the segments.

________________________________________
From: Nicholas Polys <npolys at vt.edu>
Sent: February 3, 2017 9:53 AM
To: doug sanden
Cc: x3d-public at web3d.org
Subject: Re: [x3d-public] simulating surgery > realtime plastics > cuttable, deformable

we kind of started... note also the EU project SOFA simulator



Ullrich, S., Kuhlen, T., Polys, N. F., Evestedt, D., Aratow, M., & John, N. W. (2011). Quantizing the void: extending Web3D for space-filling haptic meshes. In MMVR (pp. 670-676).

http://s3.amazonaws.com/academia.edu.documents/39382079/Quantizing_the_void_extending_Web3D_for_20151023-16794-qri8jn.pdf?AWSAccessKeyId=AKIAIWOWYYGZ2Y53UL3A&Expires=1486144363&Signature=%2ByHgyoGBtCyiCpMj6U0iC0gZbfc%3D&response-content-disposition=inline%3B%20filename%3DQuantizing_the_Void_Extending_Web3D_for.pdf

On Thu, Feb 2, 2017 at 9:22 AM, doug sanden <highaspirations at hotmail.com<mailto:highaspirations at hotmail.com>> wrote:
Q. has (something like) this already been done:

simulating surgery > realtime plastics > cuttable, deformable
- similar to volume rendering
-- separate shader for this type
-- raycasting
- unlike volume rendering
-- voxels are connected by dense set of connector edges |X =|
-- raycast to first edge-delimited triangle (surface)
-- raycasting has to figure out where surface is, expensive
-- connector edges can be stretched to deform
-- edges can be cut ie surgery cut
--- no need to create new edges/faces during cut, just destroy
--- 'healing' would re-join based on proximity
options:
- voxels just geometry - connector edges and xyz vertices - and texture xyz and texture ID
- 2-step deformation: 1) change edge lengths based on springiness/stickiness, cuts 2) recompute per-voxel xyz
- 3-step deformation: 2-step, then re-compute 2D surface for normal rendering
- outer vertices initialized to different texture ID, for surface texture
- slice textures rendered on-demand from dense models



_______________________________________________
x3d-public mailing list
x3d-public at web3d.org<mailto:x3d-public at web3d.org>
http://web3d.org/mailman/listinfo/x3d-public_web3d.org



--
Nicholas F. Polys, Ph.D.

Director of Visual Computing
Virginia Tech Research Computing

Affiliate Professor
Virginia Tech Department of Computer Science



More information about the x3d-public mailing list