[Insight-users] Levelset embedding function
Arnaud Gelas
arnaudgelas at gmail.com
Wed Feb 1 10:04:38 EST 2012
If you save an image in some term class, it is most likely normal you'd
see a difference between the resulting level set and the one you save.
Indeed term classes are accessed before the evolution, whereas the
output is the outcome of the iteration.
If your goal is to visualize or save the level set function throughout
the evolution, there are existing classes for this purpose. These
classes are in the Module LevelSetsv4Visualization (part of the
Segmentation group).
Else, I would need some more information to help...
On 02/01/2012 09:53 AM, Hui Tang wrote:
> Ha Arnaud
> Thak you for your reply.
>
> I used sparse Whitaker
> I wrote a piece of code inside itkLevelsetXXTerm
> I did it via
> const LevelSetOutputRealType value =static_cast<
> LevelSetOutputRealType >( this->m_CurrentLevelSetPointer->Evaluate(
> it.GetIndex() ) );
> //LevelSetPointer levelSet =
> this->m_LevelSetContainer->GetLevelSet( this->m_CurrentLevelSetId );
> const LevelSetOutputRealType d_val1 =
> this->m_Heaviside->Evaluate( -value );
>
> I have an iterator "it" loop over an image "m_CurreneLevelSet" which
> is initialized.
>
> I did save these images........
>
> Best,
>
> Hui
>
> On Wed, Feb 1, 2012 at 2:57 PM, Arnaud Gelas <arnaudgelas at gmail.com
> <mailto:arnaudgelas at gmail.com>> wrote:
>
> Hui,
>
> With all level set discrete representations (dense, sparse
> Whitaker, sparse Shi or sparse Malcolm), the level set function at
> the last iteration (after the evolution) should be the same as the
> output.
>
> Did you use an observer and a command to save these images? or did
> you modify some piece of code?
>
> Do you save your image at the end of one iteration (after the
> evolution gets done) ?
> How do you evaluate the level set function?
>
> Arnaud
>
>
> On 02/01/2012 08:50 AM, Hui Tang wrote:
>> ha Arnaud, thank you for reply.
>>
>> I am using ITK4.....embedding image is the signed distance image
>> whose zero value shows the current surface.
>> I runned a program that combines several itk levelset terms.
>> I would like to monitor the surface evolution, so I saved the
>> signed distance image in every n iteration.
>> but I expect the last signed distance image should be the same as
>> the final segmentation, i.e. that one that I got outside ITK
>> classes using the code from ITK example.
>>
>> Best,
>>
>> HUi
>> On Wed, Feb 1, 2012 at 2:41 PM, Arnaud Gelas
>> <arnaudgelas at gmail.com <mailto:arnaudgelas at gmail.com>> wrote:
>>
>> Hi Hui,
>>
>> It's not clear to me what you actually mean by "embedding
>> function" / "embedding image" ?
>> This function / image are embedded with respect to which class?
>> What kind of level-set representation are you using?
>>
>> Thanks,
>> Arnaud
>>
>>
>> On 02/01/2012 08:32 AM, Hui Tang wrote:
>>> Dear itk users,
>>>
>>> I outputed my intermediate embedding function every n
>>> iterations (I got it inside one of the levelset terms), but
>>> I compare that with the one that I got after evolution
>>> (level_set->GetLabelMap()->GetPixel(idx) ), they are pretty
>>> much different.
>>>
>>> Should itk embedding image in the last iteration be the same
>>> as the output embedding image?
>>>
>>> Looking forward to your reply.
>>>
>>> Best, Hui
>>>
>>>
>>>
>>>
>>> _____________________________________
>>> Powered bywww.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
>>
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20120201/ca3fc027/attachment.htm>
More information about the Insight-users
mailing list