[Insight-developers] Re: DeformableRegistration7.cxx bug report

[Luis Ibanez]

Hi Simon,

Thanks for pointing this out.

You are right, we were missing to compute the third component
of the displacement.

In fact the expression should have been written just as:

       movingPoint = transform->TransformPoint( fixedPoint );
       displacement = movingPoint - fixedPoint;
       fi.Set( displacement );


given that the itk::Point class offers the subtraction operator,
and therefore allows to write code that is dimension-independent.


The fix has been committed to the CVS repository.


Please let us know if you find any other problem.



    Thanks



       Luis



========================
Simon Warfield wrote:
> There is a bug in the displacement field computation in the example.
> The 3D field is treated as a 2D field, perhaps because the 3D example 
> was based on a 2D example.
> 
> Around line 495, the file says:
>        displacement[0] = movingPoint[0] - fixedPoint[0];
>        displacement[1] = movingPoint[1] - fixedPoint[1];
> but the displacement field in this example is in 3D.  The 
> displacement[2] should also be updated e.g.
>        for (unsigned int i = 0; i < ImageDimension; i++) {
>        displacement[i] = movingPoint[i] - fixedPoint[i];
>      }
> or by adding
>     displacement[2] = movingPoint[2] - fixedPoint[2];
>