[Insight-users] Getting segments after using watershed filter

Luis Ibanez luis.ibanez at kitware.com
Fri Jul 10 11:35:17 EDT 2009


Hi David,


                        Your lucky day !


Today, the easy way to do this is to take the output of the Watershed
filter, convert it to a LabelMap, and then use the Attribute Filters
contributed by Gaetan Lehmann in his Insight Journal paper:


    http://www.insight-journal.org/browse/publication/176


Please note that some of the filters have already been moved into
the Insight/Code/Review directory, while others are being converted
in the NAMIC Sandbox:

http://svn.na-mic.org/NAMICSandBox/trunk/ITKBinaryAttributeMorphology/


Please read Gaetan's paper to find details on how you could computes
Atributes such as "roundness" and use them to extract objects from
the labels in the image.



     Regards,


        Luis



-------------------
David Doria wrote:
> The watershed example produces a beautiful segmentation of my image, but 
> I can't figure out how to get something to the effect of a vector of 
> binary masks, one for each segment. As an example, say there was a very 
> simple image of a circle and a square. I'd want to know which pixels 
> belong to the background, which pixels belong to the circle, and which 
> pixels belong to the square.  I saw a post from 5 years ago that sounded 
> like you have to dig into the internal hash table of 
> itkWatershedSegmentTable. Is this still the case? Does anyone have an 
> example of separating the regions after the segmentation?
> 
> Thanks,
> 
> David
> 
> 
> ------------------------------------------------------------------------
> 
> _____________________________________
> 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


More information about the Insight-users mailing list