[Insight-users] fiber segmentation

Luis Ibanez luis.ibanez at kitware.com
Sat Apr 29 15:19:15 EDT 2006


Hi Esen

LeveSets may not be the best approach for segmenting
thin fibers out of images.

However, if you want to insist in LevelSets, the algorithm
that may fit your needs for segmentation of thin fibers
is the CurvesLevelSet:

http://www.itk.org/Insight/Doxygen/html/classitk_1_1CurvesLevelSetImageFilter.html


You will find an example for it in

           Insight/
              Examples/
                 Segmentation/
                      CurvesLevelSetImageFilter.cxx


This method is based on the paper:

    L. Lorigo, O. Faugeras, W.Grimson, R. Keriven, R. Kikinis
    A. Nabavi, and C.F. Westin
    "Curves: Curve evolution for vessel segmentation"
    Medical Image Analysis
    5: 195-206, 2001


You may find useful to look also at the use of the Hessian filter
such as in the applications


                  InsightApplications/
                               DuctExtractor

                  InsightApplications/
                               Curves3DExtractor



and described in

      http://www.itk.org/HTML/Curve3DExtraction.htm


Of course, you may need to convert them to 2D instead of 3D,
which shouldn't take too long, since it involves mostly
deleting code.


BTW Since you pointed out that your images are noisy,
you should explore the use of the gradient anisotropic
diffusion filters as described in the ITK Software Guide

            http://www.itk.org/ItkSoftwareGuide.pdf


These filters will reduce the noise in your images,
and will tend to preserve the details of the fibers.



   Regards,




      Luis




==========================
Seniha Esen Yuksel wrote:
> Hi all,
> 
> I am trying to segment some thin noisy fibers from histology images. Is 
> there any level set or such algorithm in ITK to do this kind of thing?
> 
> Thanks for help!
> Esen
> 

> 
> _______________________________________________
> Insight-users mailing list
> Insight-users at itk.org
> http://www.itk.org/mailman/listinfo/insight-users




More information about the Insight-users mailing list