[Insight-users] Bug in DeformableRegistration6.cxx?

[Luis Ibanez]

Hi Kristina,


Thanks for pointing this out.

You are right in all three observations.

This was entered as Bug # 1027 in the bugTracker database
http://www.itk.org/Bug/bug.php?op=show&bugid=1027&pos=26


a) We need to add

       transformHigh->SetParameters( parametersHigh );

    in order to force the internal creation of the
    coefficient images in the transform.


b) "cit" is not being incremented.
     This is now an infinity loop :-(


c) SetTransform() should use the high resolution BSpline
    transfrorm.

d) Saddly enough, there was also a bug in the CMakeLists.txt
    file, making DeformableRegistration4 run as if it was the
    test for DeformableRegistration6. Which explains why we
    didn't saw any of those problems reflected in the Dashboard.


We are committing fixes for all three points to cvs.
New comments were also inserted.


Please update your CVS checkout in order to get those
fixes and let us know if you encouter any other problem.


Our run of the example (as a test) showed an over deformed
vector field... which means that some coffee should be
invested in the procees of fine tunning the parameters of
this example before it works correctly.



Thanks



    Luis



----------------------
Kristina Grunewald wrote:

> Hello everybody,
> 
> I think there are a couple of bugs in
> /Examples/Registration/DeformableRegistration6.cxx (I am using the
> latest cvs version).
> 
> a)
> The parameters used for the second registration are not passed to the
> transform.
> line 310: transformHigh->SetParameters( parametersHigh );
> b)
> The iterator used to instantiate the parameters for the second
> registration is not counted up.
> line 349: ++cit;
> c)
> The wrong transform is used for the second registration.
> line 367: registration->SetTransform( transformHigh )
> instead of  registration->SetTransform( transformLow )
> 
> Regards,
> Kristina
> 
> 
> _______________________________________________
> Insight-users mailing list
> Insight-users at itk.org
> http://www.itk.org/mailman/listinfo/insight-users
>