[Insight-users] Narrowbanding in Geodesic Active Contour Level Set Image Filter

[Joshua Cates]

Hello Jayant,

All filters inheriting from itk::SegmentationLevelSetImageFilter use a
narrow-banding method.  There is no need to adjust the width because the
method used (sparse-field) maintains the minimum band width necessary for
the computations at all times.

Josh.

______________________________
 Josh Cates
 Scientific Computing and Imaging Institute
 University of Utah
 (801) 587-7697
 http://www.sci.utah.edu/~cates


On Wed, 19 May 2004, Jayant Chauhan wrote:

> Hey fellas,
> I seem to have got confused yet again.
> 1) Since I am just using Geodesic Active Contour (GAC) Levelset Image Fil=
ter, and not
> fastmarching, can I set the narrowbanding on and also the width. Does GAC=
 filter not use
> narrow banding ?! There seems to be no method in the itk::SegmentationLev=
elSetImageFilter
> which allows narrow banding.
> =A0
> 2) ALso, this one is a bit more on the theoretical side. As you guys alre=
ady might have
> known from my previous queries, I am working on motion estimation and tra=
cking.
> Till now I have been able to detect and track the image in the first fram=
e (using the
> difference image of the first 2 frames). Now I intend to use the output a=
s the initial
> level set for the next frame. But there is a problem. The front gets stuc=
k on other edges
> aswell thus missing out on the actual object that we had just tracked. An=
y solution
> anybody ?!!!
> =A0
> 3) Has anybody not workd on Motion estimation and tracking on a video (st=
atic camera
> assumed).
> with regards
> Yours Sincerely
> Jayant
> =A0
>
> _________________________________________________________________________=
________________________
> Earn without investing. Sell anything on www.baazee.com.
> _______________________________________________ Insight-users mailing lis=
t
> Insight-users@itk.org http://www.itk.org/mailman/listinfo/insight-users
>