[Insight-developers] Changing Transform( PhysicalPoint -> Ind ex ) methods in itk::Image

Lorensen, William E (CRD) lorensen@crd.ge.com
Tue, 26 Feb 2002 12:38:06 -0500


Luis,
Is it your plan to drive out all errors before proceeding to the next step?

Bill

-----Original Message-----
From: Luis Ibanez [mailto:luis.ibanez@kitware.com]
Sent: Tuesday, February 26, 2002 11:06 AM
To: dmshelto@andrew.cmu.edu
Cc: insight-developers@public.kitware.com
Subject: [Insight-developers] Changing Transform( PhysicalPoint -> Index
) methods in itk::Image



Hi Damion,


I'm about to change the itk::Image methods:

- TransformPhysicalPointToIndex()
- TransformPhysicalPointToContinuousIndex()
- TransformIndexToPhysicalPoint()
- TransformContinuousIndexToPhysicalPoint()


The proposed changes are the following:


1) These methods will be declared "const"
   as evaluating a point should not change the image
   and we want to be able to call these methods from
   const images.

2) Affine transforms will no longer be instantiated inside.
   Instead, a 'default' Affine transform will be set at Image
   construction time. After that, we assume that whatever
   transform is plugged into the image it is one member
   of the Transforms hierarchy and it is capable of calling
   TransfromPoint()



That could break some code temporarily...


Of special interest is to see if the new versions satisfy
what you need for supporting ultrasound images and
non-orthogonal grids...




   Luis





_______________________________________________
Insight-developers mailing list
Insight-developers@public.kitware.com
http://public.kitware.com/mailman/listinfo/insight-developers