[Insight-users] connect my own similarity measure and ITK optimization?

[Luis Ibanez]

Hi Heartfelt

The ITK registration framework was designed to
permit the easy replacement of components.

You simply need to write your own Image Metric,
make it derive from the ImageToImageMetric<> class:

http://www.itk.org/Insight/Doxygen/html/classitk_1_1ImageToImageMetric.html

and implement the GetValue() method.

You can optionally implement the GetDerivative() and
GetValueAndDerivative() if you anticipate to use your
metric with optimizers that are based on derivatives,
such as the GradientDescent optimizers.


You may want to take a look at the code of current
metrics, such as the


   MeanSquaresImageToImageMetric

or the

   CardinalityMatchImageToImageMetric


Once you define your own metric, you simply connect
it to the RegistrationMethod as it is shown in the
many examples in the ITK Software Guide


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


and their associated source code examples in


      Insight/Examples/Registration




Regards,



    Luis



-----------------------
Goo wrote:
> Hello for Insight all:
> 
> Since the requirement of research regarding medical image registration,
> a similarity measure criterion is developed by myself.
> But the other components like  optimization  or  transformation  is  
> required.
> I want to use ITK to achieve these components,but the problem in connection
> between my own  similarity measure and ITK's optimization or 
> transformation is appeared.
> 
> Can give me any suggestion?
> 
> 
> heartfelt thanks.
> 
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Insight-users mailing list
> Insight-users at itk.org
> http://www.itk.org/mailman/listinfo/insight-users