[Insight-users] Convert point correspondences to densedeformation field?

[Miller, James V (Research)]

Zachary, 

If I recall correctly, the paper referenced describes how the thin plate spline, 
the elastic body spline, and the volume spline can all be calculated 
in the same framework using slightly different kernel functions
and neighborhood relationships.

Jim

-----Original Message-----
From: insight-users-bounces+millerjv=crd.ge.com at itk.org
[mailto:insight-users-bounces+millerjv=crd.ge.com at itk.org]On Behalf Of
Zachary Pincus
Sent: Tuesday, October 11, 2005 7:48 PM
To: insight-users @ itk. org
Subject: Re: [Insight-users] Convert point correspondences to
densedeformation field?


OK, to reply to my own email, it seems that the Thin-Plate Spline  
method (available in ITK) is a common way to do this.

Several questions still stand:
(1) Is the TPS the only method available in ITK for converting point  
correspondences to a dense deformation field? If not, are there any  
known pros and cons of each?

(2) Why is the ITK class called a "thin-plate spline" when the  
documentation (http://www.itk.org/Doxygen/html/ 
classitk_1_1ThinPlateSplineKernelTransform.html ) states that it is  
an implementation based on a paper which presents the "elastic body  
spline" which is a related but not identical technique? This is  
mostly out of curiosity, but what would the correct term for the  
algorithm be? Has the "elastic body spline" term fallen into disuse  
and been grouped into the "thin-plate spline" category?

Thanks,

Zach

On Oct 11, 2005, at 3:21 PM, Zachary Pincus wrote:

> Hello folks,
>
> I'm wondering if there are any methods in ITK, or methods easily  
> implementable with tools from ITK, for converting a (very) sparse  
> deformation field (e.g. a set of point correspondences) to a  
> smooth, dense field?
>
> Perhaps a good approach would be to follow that of the  
> InverseDeformationFieldImageFilter
> ( http://www.itk.org/Doxygen/html/ 
> classitk_1_1InverseDeformationFieldImageFilter.html ) which creates  
> a kernel-based spline from the known deformation vectors, then gets  
> a dense field from the spline?
>
> Are there any other approaches, or does anyone have any tips for  
> pursuing the above approach?
>
> Thanks,
>
> Zach Pincus
>
> Program in Biomedical Informatics and Department of Biochemistry
> Stanford University School of Medicine
>
> _______________________________________________
> Insight-users mailing list
> Insight-users at itk.org
> http://www.itk.org/mailman/listinfo/insight-users
>

_______________________________________________
Insight-users mailing list
Insight-users at itk.org
http://www.itk.org/mailman/listinfo/insight-users