No subject

arrival functions (for two different initial seeds). From these the
gradient is computed, then the dot product, and finally thresholded
for values when the dot product > 0. Therefore I suspect that you will
not see any difference in the output of the CollidingFrontsImageFilter
for different seed values -- the gradient information extracted from
the arrival functions will be the same, even they are slightly offset.

For your reference, I have attached two images showing what happens
with different seed values using the FastMarchingImageFilter.

Hope this helps.

Regards, Dan

2009/2/12 Bryn Lloyd <blloyd at vision.ee.ethz.ch>:
> Hi,
>
> Does the Value at the seed points used in the CollidingFrontsImageFilter
> have any meaning or can/should it be ignored?
> http://www.itk.org/Doxygen/html/classitk_1_1CollidingFrontsImageFilter.html
>
> I have modified the CollidingFrontsImageFilterTest to use different seed
> values, but cannot find any difference in the result.
>
> In the ITK manual the value is set to zero, but no explanation is given.
>
> NodeType node;
> const double seedValue = 0.0;
> node.SetValue( seedValue );
> node.SetIndex( seedPosition );
>
> Thanks
> Bryn