[Insight-users] ITK + VTK + 3D TIFFs

[Gib Bogle]

Yes, of course.  I use the labels to count the pixels, and to see which objects 
have been detected.  I am talking about relatively large connected regions - 
with of the order of 100 pixels and roughly circular shape - that are missed. 
The connectivity setting is not relevant here.  My code performs the sequence of 
steps needed to remove the objects.

I can see nothing to discriminate between the large objects that are detected 
(labeled) and those that are missed.  If you are interested I can send you the 
image, and my code.


On 25/04/2011 7:49 p.m., Richard Beare wrote:
> Hi,
> ConnectedComponentImageFilter is only the labelling step. Each
> component comes out with a different ID, which you can then use to
> index other measurements.
>
> A filter that does the series of steps needed is
> itkBinaryShapeOpeningImageFilter (available in the review section).
>
> You need to confirm that the connectivity setting matches your
> expectation - do you want diagonal connections to count, or not. This
> is the FullyConnected option.
>
> On Mon, Apr 25, 2011 at 4:40 PM, Gib Bogle<g.bogle at auckland.ac.nz>  wrote:
>> Meantime I've been figuring out how to use itkConnectedComponentImageFilter,
>> the idea being to find all the objects then remove those below a threshold
>> size. I'm getting strange results with my small test image.
>>
>> The image is binary, 437x430.  Initially (using pixel type unsigned char) I
>> got the error message about too many objects, so I changed the type to
>> unsigned short.  The number of found objects is 838.  The image is basically
>> white, with black spots, some of which enclose white regions (these are what
>> I want to remove).  I'm finding that some of these enclosed "objects" are
>> missed.  There are four with more than 100 pixels (excluding the white
>> background), and only two are identified as objects.
>>
>> Do I have wrong expectations of itkConnectedComponentImageFilter?
>>
>> Thanks
>> Gib
>>
>> On 25/04/2011 4:42 p.m., Bill Lorensen wrote:
>>>
>>> You could use one of itk's voting filters:
>>> itkLabelVotingImageFIlter
>>> itkVotingBinaryHoleFillingImageFilter
>>> itkVotingBinaryIterativeHoleFillingImageFilter
>>>
>>>
>>> On Sun, Apr 24, 2011 at 9:28 PM, Gib Bogle<g.bogle at auckland.ac.nz
>>> <mailto:g.bogle at auckland.ac.nz>>  wrote:
>>>
>>>     I'm not sure if this should be here or on the VTK message list, so I'm
>>>     posting it to both lists.
>>>
>>>     I'm processing 3D TIFFs with ITK, but I want to use the island removal
>>>     capability of VTK, to filter each 2D image in the 3D stack (as far as I
>>> can
>>>     see VTK doesn't know about 3D TIFFs).  I presume I need to open the 3D
>>> image
>>>     in ITK, select a 2D slice, convert it to VTK 2D image format, apply the
>>>     island removal filter, convert it back to ITK image format, and write
>>> it
>>>     back into the 3D image.  It would be very nice if someone has done
>>> something
>>>     similar to this and could show me how to do it.
>>>
>>>     Thanks
>>>     Gib
>>>     _____________________________________
>>>     Powered by www.kitware.com<http://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
>>>
>>>
>> _____________________________________
>> 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
>>