[Insight-users] Segmentation validation Curves, Spheres

[Juan Cardelino]

Hi elhadj,
              In the 2D case I know some works that, compare
segmentations, by either compare boundary maps or partitions. If your
ultimate goal is segmentation, I'm not sure if it is a good idea to
compare curves, I'd rather compare segmentations. Regarding boundary
comparison you can check the work of D. Martin at Berkeley, and a
benchmark system called Berkeley Segmentation Dataset or something
like that. Regarding the region based approach, you should check the
work of J. Cardoso, search for "Quantitative evaluation".
If you really want to compare curves, there's I know some work
comparing curves represented as distance maps, which compare them
modulo a rigid transformation.
I hope this helps.
Best regards,
                  Juan

On Wed, Jul 14, 2010 at 2:20 PM, elhadj meljane
<elhadj.meljane at gmail.com> wrote:
> Hi all,
> I'm looking for a tool to validate my segmentation results in 2D and 3D:
>   - 2D: I have manual and automatic segmented curves
>   -  3D: I have a manual segmentation of a sphere-like structure and a
> simulated sphere like structure
>
> I would like to know:
>
> -  Do I have to choose manually a set of points on the manual and automatic
>    segmentation to compute a distance between the two sets (for example
>    housdorff distance)?
> - there is an itk tool (or a simplest way) that I can use it to validate the
> segmentation?
>    Thank you
>    Sincerely,
> H.
> _____________________________________
> 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.html
>
> 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
>
>