[Insight-users] label object manipulation

[Gaëtan Lehmann]

Le 28 juin 09 à 22:24, lynx.abraxas at freenet.de a écrit :

> Hello!
>
>
> Since  this  extension  seems  to  be  exactly what I'm looking for  
> I tinkered
> around a little to get familiar with itk and these filters. Taking  
> the Insight
> Journal  Article 176 - "Label object representation and manipulation  
> with ITK"
> as  documentation  I'm  wondering  if   EquivalentEllipsoidSize    
> yields   the
> ellipsoids  axes  lengths or the eigenvalues as discribed in "A  
> Label Geometry
> Image Filter for Multiple Object Measurement" (Id 301)?

The ellipsoids axes lengths.
Please be sure to use the last version, which may contain some bug  
fixes.

   http://voxel.jouy.inra.fr/darcs/contrib-itk/binaryAttributeMorphology/binaryAttributeMorphology.tar.gz

>
> I hope I haven't missed this in the discription  but  I'm   
> wondering  how  the
> filters  decide   what particles are separated. I know from MAVI  
> (labelling by
> reconstruction) that I can choose a kernel of six neighbouring   
> voxels  or  26
> (in  3D). Is there such a parameter in the itk filters? If not: What  
> kernel is
> used or is it done completly different?

It's SetFullyConnected(bool). With "true" as parameter, you get 26- 
connectivity, with "false", you get 6-connectivity.

>
> An other question I have: Can I use the  labelling  of  a  labeld   
> water  shed
> filter output as the labelling in the analysis?

You can use itk::LabelImageToLabelMapFilter to convert the label image  
produced by the watershed to a LabelMap.
itk::LabelImageToShapeLabelMapFilter and  
itk::LabelImageToStatisticsLabelMapFilter may be useful too, as they  
also compute the values of some attributes, in a single filter.

>
> A feature I'd be interested in would be a surface area analysis of  
> each label.
> As far as I understand this is connected to the surface from the  
> marching cube
> algorithm.  So  I wonder if this would be a feature of vtk but does  
> it support
> analysis of labeled objects?

You can already retrieve an estimation of the surface with this  
contribution. It's a bit longer to compute than the other attribute,  
so you have to call SetComputePerimeter(true) on the filter. Then you  
can use GetPerimeter() to get the surface of an object.

See http://voxel.jouy.inra.fr/darcs/contrib-itk/binaryAttributeMorphology/attribute_values_test.cxx 
  for an example.

Gaëtan

>
> Many thanks for the filters and the great work on itk
> Regards,
> Lynx
>
> _____________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at
> http://www.kitware.com/opensource/opensource.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

-- 
Gaëtan Lehmann
Biologie du Développement et de la Reproduction
INRA de Jouy-en-Josas (France)
tel: +33 1 34 65 29 66    fax: 01 34 65 29 09
http://voxel.jouy.inra.fr  http://www.itk.org
http://www.mandriva.org  http://www.bepo.fr

-------------- next part --------------
A non-text attachment was scrubbed...
Name: PGP.sig
Type: application/pgp-signature
Size: 203 bytes
Desc: Ceci est une signature ?lectronique PGP
URL: <http://www.itk.org/pipermail/insight-users/attachments/20090629/05489ce1/attachment.pgp>