[Insight-users] RE: Rotation origin

[Li, George (NIH/NCI)]

Thanks, Luis:

It is very helpful, making the resampling transform
the way it supposes to be.

Regards,

George


-----Original Message-----
From: Luis Ibanez [mailto:luis.ibanez at kitware.com] 
Sent: Sunday, February 13, 2005 2:15 AM
To: Li, George (NIH/NCI)
Cc: insight-users at itk.org
Subject: Re: Rotation origin



Hi George,

The center of rotation does not make part of the
parameters in the VersorRigid3DTransform.

You *MUST* pass the center of rotation explicitly
to the VersorRigid3DTransform that you use in the
resampling filter.


  You can simply do


   resamplingTransform->SetCenter(
             registrationTransform->GetCenter() );


A quick look at the source code of the class


    Insight/Code/Common/
            itkVersorRigid3DTransform.txx


will show you how the parameters are used in this
transform.  Look at lines 146-162 and lines 88-130.



   Regards,


      Luis



-----------------------------
Li, George (NIH/NCI) wrote:

> Hi, Luis and ITK users and developers:
> 
> I want to want to verify one thing about rotation origin.
> 
> I am doing the registration with Mattes metric and the 
> VersorRigid3DTransform with CenteredTransformInitializer. Prior to 
> registration, initializer->GeometryOn() is called. After the 
> registration, can the ResampleImageFilter that take the final 
> registration transformation parameters get the same rotational origin 
> as the geometric center of the image volume, as did in the 
> registration optimizer?
> 
> If not, is there a way to initialize ResampleImageFilter
> with the same origin? 
> 
> Thanks in advance for any confirmation or suggestions.
> 
> George
> 
> 
>