[Insight-users] timestep in level sets

[Luis Ibanez]

Hi Sovira,


Please take a look at the code of the FiniteDifferenceImageFilter:


        Insight/Code/Common/itkFiniteDifferenceImageFilter.txx


Note that the LevelSet filters in ITK have the following (partial)
hierarchy:

http://www.itk.org/Insight/Doxygen/html/classitk_1_1SegmentationLevelSetImageFilter.html

They all derive from the "SegmentationLevelSetImageFilter" class,
that in turns has these superclasses:

http://www.itk.org/Insight/Doxygen/html/classitk_1_1SparseFieldLevelSetImageFilter.html

and the

http://www.itk.org/Insight/Doxygen/html/classitk_1_1FiniteDifferenceImageFilter.html


The time steps are re-computed at every iteration based on the width of 
the narrow band used around the zero-set.  The method that computes the
time-step in every thread is "CalculateChange()", and the method that
chooses the final step for the iteration by taking all the time-steps
computed by all the threads is "ResolveTimeStep()".  The first method is
overloaded in classes derived from the FiniteDifferenceImageFilter,
while the second method is implemented in the file 
itkFiniteDifferenceImageFilter.txx in lines 168 - 199.



     Regards,



         Luis



-----------------------------------
Tan, Sovira (NIH/CC/DRD) [F] wrote:
> Hi,
> 
>  
> 
> I would to know how the timestep is computed in the ITK implementation 
> of level sets. In particular I’m interested in the equation used in 
> computing the timestep and I was wondering if it were computed once at 
> the beginning or if there’s a new timestep at each iteration.
> 
>  
> 
> Thank you,
> 
> Sovira
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Insight-users mailing list
> Insight-users at itk.org
> http://www.itk.org/mailman/listinfo/insight-users