[Insight-users] Using Example - MutualInformationEuler2DRegistration..

Luis Ibanez luis . ibanez at kitware . com
Mon, 12 Aug 2002 16:35:28 -0400


Hi Venkat,

1) In order to see the Registered image you have to go to the
    "Display" frame on the GUI and select "RegisteredImage"
then click on the "Show" button just below.  The registered
image should popup in an independent window.

You can also get a superposition of the fixed image and the
mapped moving image by selecting "MixedChannelView" and then
clicking on the "Show" button.  The fixed and moving images
will be display using different RGB channels. You can modify
the channel assignment in the "Advanced Options..." menu.


2) This example has already been extended for several transforms
    including Affine. The corresponding classes are:

- MIScaleRegistrato.h         : ScaleTransform
- MIAffineRegistrator.h       : AffineTransform
- MISimilarity2DRegistrator.h : Similarity2DTransform

Just go to the RegisterApplication.h file and comment/uncomment
the lines 33-37.  These lines have the typedef defining which
Transformtion class to use for the registration process.
Please, make sure to uncomment only one of them.


3) The support for PNG images in MutualInformationEuler2DRegistration
is visible in the code at:

   RegisterApplication.cxx :line 50  where a PNGImageIO factory is registed:

   itk::PNGImageIOFactory::RegisterOneFactory();

Just select any .png file when you are in the file dialog.

--

Please let us know if you encounter any further problems


Thanks



   Luis

===================================================

Venkat Anand wrote:
> Hi,
> 
> I am trying to rerun the example - MutualInformationEuler2DRegistration.
> I used two simple images (A black filled circle on a white background at a
> slightly offset position in the moving imag as compared to the original
> image format and couldnt find any change in the registered image  at
> all.) 
> 
> 1. Looking at the code, i do see the support for PNG images..? any
> pointers?
> 
> 2. I tried to rework the code for Affine transformation..(the code given
> in the example was for rigid transformation) and still couldnt find any
> improvements.
> 
> 
> Any help appreciated.
> 
> -- Venkat
> 
> ******************************
> Venkatanand Venkatachalapathy
> 
> ph: (H) 801 533 0387 
> http://www.cs.utah.edu/~venkat
> ******************************        
> 
> 
> 
> 
> _______________________________________________
> Insight-users mailing list
> Insight-users@public.kitware.com
> http://public.kitware.com/mailman/listinfo/insight-users
> 
>