[Insight-users] itk::GeodesicActiveContourLevelSetImageFilter

[Vaaksiainen]

Hi,

0-1 image doesn't sound distance image to me. Try

binary <-- binary - 0.5,

so that your image is negative inside and positive outside (need not
be distance image). If the input to the filter contains wide areas of
zero-levelset (e.g. the entire inside), the evolution won't work (at
least the way expected). Also, the curvature values are scaled by the
feature image, so if that is zeros, non will happen. Otherwise, it
should converge towards N-D sphere.

Best,

-Vaaksiainen

2012/12/19 Suyash P. Awate <suyash at cs.utah.edu>:
>
> Hi,
>
> I'm trying to use the itk::GeodesicActiveContourLevelSetImageFilter filter.
>
> I pass to SetInput() a signed-distance-transform image (inside 0, outside 1)
> where the zero level set is close to the boundary of the structure  desire
> to segment.
>
> Even if I don't use any advection and propagation terms (
> SetAdvectionScaling(0), SetPropagationScaling(0) ) and just use a smoothness
> term ( SetCurvatureScaling(1) ), I get a terribly-wiggly zero level set in
> the output image. Am I missing something concerning this ITK-filter usage ?
> [ my image spacing is isotropic ].
>
> Thanks for your time.
>
> Regards,
> S.
>
>
> _____________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at
> http://www.kitware.com/opensource/opensource.html
>
> Kitware offers ITK Training Courses, for more information visit:
> http://www.kitware.com/products/protraining.php
>
> Please keep messages on-topic and check the ITK FAQ at:
> http://www.itk.org/Wiki/ITK_FAQ
>
> Follow this link to subscribe/unsubscribe:
> http://www.itk.org/mailman/listinfo/insight-users
>