[Insight-users] automatic segmentation

[Dawood Masslawi]

I haven't worked with those tools before but if the results satisfy your application
requirements there is no harm in using them. 
Regards,
Dawood

--- On Tue, 11/9/10, H-B <h4cd at yahoo.com> wrote:

From: H-B <h4cd at yahoo.com>
Subject: Re: [Insight-users] automatic segmentation
To: "Dawood Masslawi" <masslawi at gmail.com>, "Jon Haitz Legarreta Gorroño" <jhlegarreta at vicomtech.org>
Cc: insight-users at itk.org
Date: Tuesday, November 9, 2010, 10:50 AM

Dawood and Jon, i really thank you about your interest in this subject
 
i tautly agree with you, if there is no good tool i will implement it my self.
 
to save time, did you try 3D Slicer or Seg3d.
i found those tools while i was searching.
 
do you advice me to use it or implement segmentation my self?
 
Best Regards
 
  H.B

--- On Mon, 11/8/10, Jon Haitz Legarreta Gorroño <jhlegarreta at vicomtech.org> wrote:


From: Jon Haitz Legarreta Gorroño <jhlegarreta at vicomtech.org>
Subject: Re: [Insight-users] automatic segmentation
To: "Dawood Masslawi" <masslawi at gmail.com>
Cc: insight-users at itk.org
Date: Monday, November 8, 2010, 8:17 AM


Hi there,
reading Dawood's mail I just realized that I (possibly we) forgot to make some little but important comments on the issue.

IMHO, there are (at least) three main difficulties when developping a segmentation method and/or using an already implemented segmentation method:

1. Good parametrization. A good selection of the parameters is crucial for the method.

2. The parameters' value is highly dependent on the image type and region to be segmented. Sometimes some compromise values work, but trial and error is usually necessary to find the appropriate range of values for your application.

3. Automatization of the method. Automating the segmentation process is by no means a trivial task; it may well happen that you are not able to achieve a fully automated method (in fact, automatic segmentation is a recurrent topic in medical image processing); if the method
 performs well, a minimal user interaction is then acceptable.


HTH.

Best regards,
JON HAITZ




El 08/11/2010 13:43, Dawood Masslawi escribió: 




Dear H. B.,Basically there are four types of segmentation methods available inITK which are morphological, level-set, deformable and hybrid segmentation methods. I completely agree with Jon, the ITK software guide is a comprehensive source for learning about these different methods but considering your application (image guided surgery with FEM) any of the level-set methods seem suitable (e.g. Canny-Edge Level Set Segmentation). These methods would give you a good control over your resulting images since they are parametric
 approaches. Another good resource for learning more about segmentation methods is the book "image Processing, Analysis, and Computer Vision" by Milan Sonka et al., chapter 7 of the book is about different segmentation methods.Regards,Dawood Masslawi>Hi all,> 
>Is there a good atomatic segmentation tool, that i give it set of images and the output is set of segmented images.
> 
>Best Regards
>   H.B

_____________________________________
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



-----Inline Attachment Follows-----


_____________________________________
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



      


      
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20101109/a0a13398/attachment.htm>