[Insight-users] Medianfilter:Requested region is outside..

doege at mi.fu-berlin.de doege at mi.fu-berlin.de
Thu Mar 19 05:38:38 EDT 2009 

I did post my problem to the mitk mailinglist but didn´t get a  
response, and I only now know for sure that this is an mitk problem.

But thanks a lot for your help!
Best regards,
Juliane

Quoting Andreas Schuh <andreas.schuh.84 at googlemail.com>:

> If you have a look at the MITK docu of the Adaptor Classes and the
> macro definition of AccessByItk, you see that you need a
>
> InstantiateAccessFunction(AutomaticSegmentation);
>
> in your source file. Moreover, your access function misses the second
> parameter.
>
> template< class TPixel, unsigned int VDimension >
> void AutomaticSegmentation( itk::Image<TPixel, VDimension>* itkImage,
> TPixel* dummy = NULL )
> {
>   // ...
> }
>
> The dummy variable may ensure that the right template instantiation is used.
>
> However, you should post your problem to an MITK mailinglist if one
> exists, as the problem doesn't come from ITK.
>
> --
> regards
> Andreas
>
> Am 18.03.2009 um 15:37 schrieb doege at mi.fu-berlin.de:
>
>> Hi,
>>
>> 1) I put a try-catch-block around the update
>> 2) I tried to call itkImage->Print() and   
>> itkImage->DisconnectPipeline() but I couldn´t call the function. It  
>>  only worked
>> before I called the function AutomaticSegmentation, which is a   
>> template and I some how need it for the itk-stuff. So I guess it   
>> has something to do  with calling the template.
>>
>>
>> void QmitkAutomaticSegmentation::ImageSelected(const   
>> mitk::DataTreeFilter::Item* item)
>> {
>>   //mitk::Image* image; ( header file)
>>   image = dynamic_cast<mitk::Image*>( item->GetNode()->GetData( ));
>> }
>> void QmitkAutomaticSegmentation::StartButtonClicked()
>> {
>>
>> ..
>> image->DisconnectPipeline();
>> image->Print();
>> AccessByItk( image, AutomaticSegmentation );
>> }
>>
>> Greetings,
>> Juliane
>>
>>
>> Quoting Andreas Schuh <andreas.schuh.84 at googlemail.com>:
>>
>>> Hi!
>>>
>>> Ok, actually it would be no good behavior of itk::MedianImageFilter (or
>>> the used itk::NeighborhoodIterator) if it couldn't handle this degraded
>>> case properly.
>>>
>>> Did you try to embrace the call of the Update() method of the median
>>> filter with a try-catch-block to ensure that the problem is this update?
>>> You could print the state of the image regions just before calling
>>> update or even in the catch block. To do so use itkImage->Print().
>>> As you tried UpdateLargestPossibleRegion() already, the requested
>>> region should actually be set to the largest possible region before the
>>> filter is updated.
>>> However, if your itkImage that is passed to your
>>> AutomaticSegmentation() method is connected to an output of a previous
>>> filter, this filter
>>> may be updated as well and modify the requested region. Where does your
>>> itkImage come from?
>>> You could try to disconnect it manually, calling
>>> itkImage->DisconnectPipeline(), just to track back the error.
>>> For instance, call DisconnectPipeline() directly before passing the
>>> image to your method.
>>>
>>> Concluding, my hints to debug the error are:
>>>
>>> 1. Enclose the update call of the median filter in a try-catch block
>>> (catch any itk::ExceptionObject and print the message)
>>> 2. Print the state of the image right before the update call and within
>>> the catch block using Print()
>>> 3. Disconnect the input image manually from any previous ITK pipeline
>>> using DisconnectPipeline()
>>>
>>>
>>> BTW: Your comment "aufrufen des Filters" seems to be accidentally
>>> placed on the wrong line ...
>>>
>>> --
>>> regards
>>> Andreas
>>>
>>> doege at mi.fu-berlin.de schrieb:
>>>> Sorry, of course I doesn´t want to set it 0,0,0 it was just a try  
>>>>   after 1,1,1 didn´t work out, and I forgot to set it back, but I  
>>>>  got  the same error.
>>>>
>>>> Quoting Andreas Schuh <andreas.schuh.84 at googlemail.com>:
>>>>
>>>>> Hi Juliane,
>>>>>
>>>>> does it make sense that you want to apply the median filter without
>>>>> considering any neighborhood?
>>>>> What do you think is the median of just a single value?
>>>>> That the radius is set to [0, 0, 0] may be your problem.
>>>>> Try at least [1, 1, 1] instead, which is the default radius used by
>>>>> itk::MedianImageFilter.
>>>>>
>>>>> -- 
>>>>> regards
>>>>> Andreas
>>>>>
>>>>> doege at mi.fu-berlin.de schrieb:
>>>>>> Hi Luis,
>>>>>>
>>>>>> 1. perhaps I was misleading words, I trying to use the    
>>>>>> itk::median  filter with mitk and I think that I am getting an   
>>>>>>  itk error, which  probably means that I am using either itk or  
>>>>>>   mitk wrong.
>>>>>>
>>>>>> 2. I am getting the error after using it the first time.
>>>>>>
>>>>>> 3.:
>>>>>> template < typename TPixel, unsigned int VImageDimension >
>>>>>> void QmitkAutomaticSegmentation::AutomaticSegmentation(    
>>>>>> itk::Image<  TPixel, VImageDimension >* itkImage )
>>>>>> {
>>>>>> typedef itk::Image< TPixel, VImageDimension > InputImageType;    
>>>>>> //  Image type
>>>>>> typedef itk::Image< TPixel, VImageDimension > OutputImageType;
>>>>>> typedef typename InputImageType::IndexType    IndexType;
>>>>>>
>>>>>> //Median Filter
>>>>>> typedef itk::MedianImageFilter<InputImageType, OutputImageType   
>>>>>> >   FilterType; //creating Filtertype
>>>>>> FilterType::Pointer filter = FilterType::New(); //filter object  
>>>>>>    erzeugt filter
>>>>>>
>>>>>> InputImageType::SizeType indexRadius; //Radius for calculating   
>>>>>>  the  median filter
>>>>>>
>>>>>> indexRadius[0] = 0; // radius along x
>>>>>> indexRadius[1] = 0; // radius along y
>>>>>> indexRadius[2] = 0; // radius along z
>>>>>> filter->SetRadius( indexRadius );
>>>>>>
>>>>>>
>>>>>> filter->SetInput( itkImage); //aufrufen des Filters
>>>>>>
>>>>>> std::cout << "Execute the filter";
>>>>>> filter->UpdateLargestPossibleRegion();
>>>>>>
>>>>>> //filter->Update();
>>>>>>
>>>>>>
>>>>>> mitk::Image::Pointer resultImage = mitk::ImportItkImage(     
>>>>>> filter->GetOutput() );
>>>>>> mitk::DataTreeNode::Pointer newNode = mitk::DataTreeNode::New();
>>>>>> newNode->SetData( resultImage );
>>>>>>
>>>>>>
>>>>>> mitk::DataTreeNodeFactory::SetDefaultImageProperties( newNode );
>>>>>>
>>>>>>
>>>>>> // add result to data tree
>>>>>> mitk::DataStorage::GetInstance()->Add( newNode );
>>>>>>
>>>>>> mitk::RenderingManager::GetInstance()->RequestUpdateAll();
>>>>>> }
>>>>>>
>>>>>> Best regards
>>>>>> Juliane
>>>>>>
>>>>>> Quoting Luis Ibanez <luis.ibanez at kitware.com>:
>>>>>>
>>>>>>>
>>>>>>> Hi Juliane,
>>>>>>>
>>>>>>>
>>>>>>> 1) Why are you implementing a median filter ?
>>>>>>>
>>>>>>> Is there are reason why the current MedianFilter in ITK
>>>>>>> is not suitable for what you are doing ?
>>>>>>>
>>>>>>>
>>>>>>> 2) Do you get that error message after running the
>>>>>>> filter multiple times ?
>>>>>>>
>>>>>>> or do you get that error in the first run of the filter  ?
>>>>>>>
>>>>>>>
>>>>>>> 3) Can you please post to the list the code of your filter ?
>>>>>>>
>>>>>>>
>>>>>>> Thanks
>>>>>>>
>>>>>>>
>>>>>>>    Luis
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> -----------------------------
>>>>>>> doege at mi.fu-berlin.de wrote:
>>>>>>>> Hello,
>>>>>>>>
>>>>>>>> I am trying to implement a median filter with mitk and   
>>>>>>>> getting   the   following itk error:
>>>>>>>> ..\..\..\source\Code\Common\itkDataObject.cxx:
>>>>>>>> 397:Requested region is (at least partially) outside the    
>>>>>>>> largest   possible region.
>>>>>>>>
>>>>>>>> I tried to update only the largest possible region but it    
>>>>>>>> didn´t   work  either. Does anyone got another idea? I also    
>>>>>>>> couldn´t find   out how to  print out the size of the region.
>>>>>>>> I am using dicom images with more than 4 slices.
>>>>>>>>
>>>>>>>> Best regards
>>>>>>>> Juliane
>>>>>>>>
>>>>>>>> ----------------------------------------------------------------
>>>>>>>> This message was sent using IMP, the Internet Messaging Program.
>>>>>>>>
>>>>>>>> _____________________________________
>>>>>>>> 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
>>>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> ----------------------------------------------------------------
>>>>>> This message was sent using IMP, the Internet Messaging Program.
>>>>>>
>>>>>> _____________________________________
>>>>>> 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
>>>>>
>>>>> _____________________________________
>>>>> 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
>>>>
>>>>
>>>>
>>>> ----------------------------------------------------------------
>>>> This message was sent using IMP, the Internet Messaging Program.
>>>>
>>>> _____________________________________
>>>> 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
>>
>>
>>
>> ----------------------------------------------------------------
>> This message was sent using IMP, the Internet Messaging Program.
>>



----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.

More information about the Insight-users mailing list