[ITK] ScaleVersor3DTransform

[Yaniv, Ziv Rafael (NIH/NLM/LHC) [C]]

Hi Tobias,

Unfortunately the name of this transformation type is misleading, it is
not what you expect, it adds a diagonal scale-1 matrix to the rotation
matrix. I remember being surprised by this too but, in this rare case ;),
the documentation says what it does:


https://itk.org/Doxygen/html/classitk_1_1ScaleVersor3DTransform.html

      regards
          Ziv

p.s. A description of most ITK transformation types is available in our
SimpleITK notebooks repository
(https://github.com/InsightSoftwareConsortium/SimpleITK-Notebooks/blob/mast
er/Python/22_Transforms.ipynb). It illustrates the behavior of ITK
transformations, some of which may be unexpected.


On 11/1/16, 5:48 AM, "Wood, Tobias" <tobias.wood at kcl.ac.uk> wrote:

>Hello,
>
>I tried to create a ScaleVersor3DTransform that applies an isotropic 0.1
>scaling and a 180 degree rotation about Z. I am hence a bit confused
>about how I get the following output when I print the Transform:
>
>ScaleVersor3DTransform (0x7f8a726789b0)
>  RTTI typeinfo:   itk::ScaleVersor3DTransform<double>
>  Reference Count: 1
>  Modified Time: 209
>  Debug: Off
>  Object Name: 
>  Observers: 
>    none
>  Matrix: 
>    -1.9 0 0 
>    0 -1.9 0 
>    0 0 0.1 
>  Offset: [0, 0, 0]
>  Center: [0, 0, 0]
>  Translation: [0, 0, 0]
>  Inverse: 
>    -0.526316 0 0 
>    0 -0.526316 0 
>    0 0 10 
>  Singular: 0
>  Versor: [ 0, 0, 1, 0 ]
>  Scales:       [0.1, 0.1, 0.1]
>
>Not that the Scales and Versor parameters appear correct. However the
>Matrix (and Inverse) look very badly wrong. I think the Matrix should be:
>-0.1 0 0
>0 -0.1 0
>0 0 -0.1
>
>What have I misunderstood?
>
>Thanks,
>Toby
>
>_______________________________________________
>Community mailing list
>Community at itk.org
>http://public.kitware.com/mailman/listinfo/community