Hi Ido,
this script takes an image (actually, a feature image, i.e. the output of
vmtkimagefeatures) and a surface in input,
and moves the vertices of the surface so that they converge on the minimum of
the potential defined by the image.
For instance, you can extract an isosurface from the image and let it converge
to the ridges of the gradient magnitude
of the image, like with level sets (this is just one use case, you can of
course start from any surface, not necessarily
an isosurface).
Here is the pipe to do it:
vmtkimagereader -ifile foo.vti --pipe vmtkmarchingcubes -l 200.0 --pipe
vmtkimagefeatures --pipe vmtkpotentialfit -potentialweight 1 -curvatureweight
0.1 -iterations 100 -ofile foo.vtp
Here potentialweight is the analogous of level set advection.
Hope this will get you started.
Luca
On Jul 31, 2011, at 12:00 PM, ido yerushalmy wrote:
> Hi,
> I'm having a similar problem and want to try using vmtkpotentialfit.
> I tried using it the following way:
>
> vmtkpotentialfit -imagefile foo.vti -ofile foo_potentialfit.vtk
>
> but it seems that the output is empty (no file is created). If I try to view
> the surface, I see nothing, just a black screen.
> Does anyone know if there's an example I can follow for this script?
>
> Regards,
> Ido
>
> On Fri, Jul 29, 2011 at 11:28 AM, Luca Antiga <luca.ant...@gmail.com> wrote:
> Hi Yap,
> if you are on a 64 bit system, and assuming you compiled vmtk from scratch
> (correct me if I'm wrong)
> make sure that vmtk and its dependencies are compiled for a 64 bit
> architecture (amd64 or x86_64).
> You should find a reference to these two in your CMakeCache.txt in the build
> directories.
> If you are running on a 64 bit system, your build should already be 64 bits
> anyway...
>
> In any case, there is a possibility that memory is not sufficient if you are
> running, for instance, level
> sets on very large images. Running level sets means that your image has to be
> replicated several
> times in memory (to compute gradients, etc), and converted to "expensive"
> numerical representations,
> so it's totally possible that you run out of memory.
>
> Just so you know, you can for instance segment individual VOIs and merge them
> after segmentation,
> or you can work on a subsampled image and then optimize your segmentation on
> the full image.
> Or you could rely on less expensive deformable models - an explicit
> deformable model such as the
> one provided by vmtkpotentialfit. Or a bit of all the above.
>
> What's the right way to go really depends on the problem at hand and what
> you're trying to achieve.
>
> What is the size of your images and what algorithms are you trying to run on
> them?
>
> Luca
>
>
> On Jul 21, 2011, at 9:05 PM, Yap, Choon Hwai wrote:
>
> > Hi,
> >
> > I would like to segment very high resolution images, and keep encountering
> > memory error as such:
> >
> > ERROR: In /build/buildd/vtk-5.4.2/Common/vtkDataArrayTemplate.txx, line 141
> > vtkFloatArray (0x1d582b0): Unable to allocate 186212691 elements of size 4
> > bytes.
> >
> > Is there a way to increase the memory allocation for vmtk?
> > I'm using Ubuntu v11.04
> >
> > Any advice will be greatly appreciated..!
> > Thanks!
> >
> > regards,
> > Yap
> >
> > ------------------------------------------------------------------------------
> > 5 Ways to Improve & Secure Unified Communications
> > Unified Communications promises greater efficiencies for business. UC can
> > improve internal communications as well as offer faster, more efficient ways
> > to interact with customers and streamline customer service. Learn more!
> > http://www.accelacomm.com/jaw/sfnl/114/51426253/
> > _______________________________________________
> > vmtk-users mailing list
> > vmtk-users@lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/vmtk-users
>
>
> ------------------------------------------------------------------------------
> Got Input? Slashdot Needs You.
> Take our quick survey online. Come on, we don't ask for help often.
> Plus, you'll get a chance to win $100 to spend on ThinkGeek.
> http://p.sf.net/sfu/slashdot-survey
> _______________________________________________
> vmtk-users mailing list
> vmtk-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/vmtk-users
>
------------------------------------------------------------------------------
BlackBerry® DevCon Americas, Oct. 18-20, San Francisco, CA
The must-attend event for mobile developers. Connect with experts.
Get tools for creating Super Apps. See the latest technologies.
Sessions, hands-on labs, demos & much more. Register early & save!
http://p.sf.net/sfu/rim-blackberry-1
_______________________________________________
vmtk-users mailing list
vmtk-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/vmtk-users
