[ITK] Landmark projection onto volumetric data.

Dan Mueller dan.muel at gmail.com
Mon Jun 2 17:09:29 EDT 2014


Hi Ryan,

If I understand your problem correctly, then the Fast Marching Image Filter
may help.

Your speed function would be 1.0 outside your volume and 0.0 at the surface
and inside your volume. You could then initialize trial points as all the
points on the surface of the volume. The result of the filter will be a
time map (aka arrival function) of the volume surface expanding along its
normal until it reaches your points of interest.

You could then back project your points to the initial surface, using this
paper for example:
http://www.insight-journal.org/browse/publication/213

Hope this helps,

Cheers Dan
On 3 Jun 2014 04:23, "Ryan Edward Young" <Ryan.E.Young at colorado.edu> wrote:

> Hello,
>
> I have an algorithmic question.  I am tryping to porject points down onto
> a volume.  This is typical done by converting the volume into a mesh.  The
> the normal of the surface and the surface itself are easily accessible.
> Then the point can be project along the normal until it reaches the
> surface.  My data is volumetric and it would be best if I did not have to
> create meshes.  Any pointers on where to begin.  How would you define the
> normal?
>
> Thanks
> Ryan
>
> _______________________________________________
> Community mailing list
> Community at itk.org
> http://public.kitware.com/cgi-bin/mailman/listinfo/community
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/community/attachments/20140603/68bcac11/attachment-0002.html>


More information about the Community mailing list