[Insight-users] Watershed and Ultimate Erode Points on 3D MicroCT

[Bradley Lowekamp]

Hello,

The classic gray scale watershed segmentation takes a gradient magnitude as an input, where the binary watershed takes a distance map. I put the filters together for the binary watershed method recently with SimpleITK. The basics of the SimpleITK python code was this:

l = sitk.BinaryThreshold( labelImage, labelID, labelID, 1, 0 )

d = sitk.SignedMaurerDistanceMap( l,
                                      inInsideIsPositive = False,
                                      inSquaredDistance = False,
                                      inUseImageSpacing = True )
d = sitk.Threshold( d, -1e23, 0, 0 )

ws = sitk.MorphologicalWatershed( d, inMarkWatershedLine=True, inLevel = level )


It should be fairly straight forward to use ITK proper and create a pipeline based on the above.

Regarding the Ultimate Erode points. I don't believe we have a direct morphological filter to compute that. But it can be simply computed as the RegionalMaxima of a distance field from you binary image.

Brad

On Nov 6, 2012, at 3:11 PM, Ruben Neves <ruben.m.neves at gmail.com> wrote:

> Greetings,
> 
> I have a 16-bit (signed short) vtk image of a MicroCT that I have already filtered and binarized.
> 
> I have been playing with some samples (taken from the original) in ImageJ and I reached the point were I really need to apply the Ultimate Erode Points and the Watershed segmentation to my binary original. To be clear, I want the separate outputs from each process, no mixes...
> 
> As for the watershed segmentation, I've got the algorithm taken from the examples of ITK, but I'm not very sure how to work with the input parameters. In my binary original I have shapes that are "overlapping", and became a single shape with the binary threshold, hence my need for watershed. I just need to separate them like ImageJ does.
> 
> As for the Ultimate Erode Points, I'm a bit clueless...
> 
> Can Anyone shed some light over these two issues?
> 
> Thanks in advance, 
> Best regards
> 
> Ruben Neves
> _____________________________________
> 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.php
> 
> 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

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20121107/1c68371d/attachment.htm>