[Insight-users] region growing segmentation -- isolatedConnectedImageFilter

[Luis Ibanez]

Hi Sara,

Please note that the filter
http://www.itk.org/Insight/Doxygen/html/classitk_1_1IsolatedConnectedImageFilter.html

has two modes of operation.

A) One mode is to segment a dark region by preventing it from
    entering a neighbor bright region.

                  FindUpperThresholdOn()

B) Another mode is to segment a bright region by preventing
    it from entering a neighbor dark region.

                  FindUpperThresholdOff()

So, you have to select the proper mode, that corresponds to
the bright/dark or dark/bright combination of your case.



Note that something similar can also be done with the filter
http://www.itk.org/Insight/Doxygen/html/classitk_1_1IsolatedWatershedImageFilter.html
(although a derivative pre-processing may be needed)



     Regards,






---------------------
sara al kady wrote:
> Thanks for your reply,
> 
> But Actually I wanted to know which seeds to pick ?
> 
> I mean in the example provided with the ITKuserguide .. it is applied on 
> BrainProtonDensitySlice.png image and it uses the first seed (61,140 ) 
> and the other seed  which is from outside the white matter at ( 63,43 )
> but when I changed the points to follow the same rule - connected to 
> first seend and NOT connected to the other - so I picked (110, 160 ) and 
> the other seed also ( 63, 43) with lower intenisty 166 .. it gave 
> completely strange output.
> 
> so .. any help will be appreciated.
> 
> Thanks a lot.
> 
> ------------------------------------------------------------------------
> Date: Fri, 6 Mar 2009 14:00:06 -0500
> Subject: Re: [Insight-users] region growing segmentation -- 
> isolatedConnectedImageFilter
> From: harvey.cline at kitware.com
> To: sara_alkady at hotmail.com
> CC: insight-users at itk.org
> 
> Hi
>  
> In the example of segmenting the brain from the scalp. The first seed is 
> placed in the brain and the second seed in the scalp. The program uses a 
> binary search to find the smallest threshold value that automatically 
> separates the brain from the scalp.
> 
> 2009/3/6 sara al kady <sara_alkady at hotmail.com 
> <mailto:sara_alkady at hotmail.com>>
> 
>     hello,
> 
>     We've been trying to use region growing techniques to separate the
>     brain components-we're working on axial slices- but there is
>     something that I don't understand in the
>     IsolatedConnectedImageFilter which is .. Which 2 seeds do I have to
>     pick to separate the components ?
>     I understand that the first seed should be related to the component
>     I want and the 2nd seed from outside it, but it doesn't work with
>     any seed , I think it has to be a specific seed to make the
>     segmentation work correctly.
> 
>     if anyone knows how to pick the seed or if I misunderstood something
>     and need to use another filter completely kindly tell me.
> 
>     Thanks,
>     Sara Al-Kady
> 
>     ------------------------------------------------------------------------
>     check out the rest of the Windows Live™. More than mail–Windows
>     Live™ goes way beyond your inbox. More than messages
>     <http://www.microsoft.com/windows/windowslive/>
> 
>     _____________________________________
>     Powered by www.kitware.com <http://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
> 
> 
> 
> ------------------------------------------------------------------------
> Invite your mail contacts to join your friends list with Windows Live 
> Spaces. It's easy! Try it! 
> <http://spaces.live.com/spacesapi.aspx?wx_action=create&wx_url=/friends.aspx&mkt=en-us>
> 
> 
> ------------------------------------------------------------------------
> 
> _____________________________________
> 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