[Insight-users] question ? [ITK]

[Luis Ibanez]

Hi Agata,

You are not seeing errors because you are mapping
the very same input landmarks, and because in its
default setting the TPS transform is an interpolating
spline (which means that it actually passes by the
landmark points.

This applies to all the Transforms deriving from:
http://www.itk.org/Doxygen/html/classitk_1_1KernelTransform.html

You can change the behavior of the TPS from an Interpolating
spline to an Approximating Spline by changing the value of the
Stiffness parameter with the method:

http://www.itk.org/Doxygen/html/classitk_1_1KernelTransform.html#a82a5f8a507dd4a8bc9c55972722f9f34


" Stiffness of the spline. A stiffness of zero results in the standard
interpolating spline. A non-zero stiffness allows the spline to approximate
rather than interpolate the landmarks. Stiffness values are usually rather
small, typically in the range of 0.001 to 0.1. "

When you use it as an approximating spline, the deformation
will be smoother, and the target points of the input landmarks
will no longer coincide with the target landmarks.


    HTH


        Luis


-----------------------------------------------------------------------
On Mon, Aug 20, 2012 at 2:01 PM, Agata Krasoń <agatakrason at gmail.com> wrote:

> Hi Luis,
>
> Thank You very much for explain this area .
> Thanks Luis !
> But If I have position of points after registration (after transformation
> TPS) I don't need exacly all of matrices  and components of TPS.
> But I not sure if these positions of point are good after TPS
> transformation. Is there anything with errors ?
> When I was trying test rigid or affine. I received always values weighted
>  small errors (about 0,1 - 0,5 between measured values and registered
> values).
> Now with TPS - anything with errors. I don't see error between values.
> (all the same) ???
>
>
>
> Regards,
> Agata
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20120820/c17918e5/attachment.htm>