[Rtk-users] projection geometry offset

[Simon Rit]

Hi,
I guess the confusion comes from "“The detector position is defined with
respect to the source with three parameters”. I guess we wrote this due to
the use of source distances in the third coordinate
SourceToIsocenterDistance-SourceToDetectorDistance,
but it's with respect to the isocenter indeed. Any amendment to the
documentation is welcomed (via a github PR preferably).
In general, I strongly suggest to play with the python wrappings to
understand what positions are taken by the source and a detector coordinate
with the methods GetSourcePosition
<http://www.openrtk.org/Doxygen/classrtk_1_1ThreeDCircularProjectionGeometry.html#a8251a17e5ab907935da8b734ff4fb798>
and GetProjectionCoordinatesToFixedSystemMatrix
<http://www.openrtk.org/Doxygen/classrtk_1_1ThreeDCircularProjectionGeometry.html#ae072bdb4511a44615fb12ac48538be79*>
.
Simon

On Thu, May 12, 2022 at 2:08 PM Moritz Schaar <schaar at imt.uni-luebeck.de>
wrote:

> Dear RTK users,
>
>
>
> I am struggling with understanding RTK‘s projection geometry.
>
> My measurement setup is that source and detector are fixed and the object
> is on a rotating plate.
>
> The rotation axis has a slight horizontal offset as it can be seen in
> projections measured with 180° angular shift.
>
>
>
> My assumptions based on http://www.openrtk.org/Doxygen/DocGeo3D.html now
> are:
>
> 1)      The isocenter is fix and in my case the center of the rotation
> axis of the plate.
>
> 2)      The documentation says that “The source position is defined with
> respect to the isocenter with three parameters”.
>
> So I should apply the estimated horizontal offset as SourceOffsetX.
>
> 3)      The documentation also says that “The detector position is
> defined with respect to the source with three parameters”
>
> Does this imply that SourceOffsetX also has an impact on the detector
> position, e.g. the detector origin (0,0)?
>
> Because this would be contrary to the next sentence of the documentation:
>
> “(ProjectionOffsetX,ProjectionOffsetY,SourceToIsocenterDistance-SourceToDetectorDistance)
> are the coordinates of the detector origin (0,0) in the rotated coordinated
> system.”
>
> So either the detector is also positioned with respect to the isocenter,
> then the coordinates are
> (ProjectionOffsetX,ProjectionOffsetY,SourceToIsocenterDistance-SourceToDetectorDistance)
>
> or respect to the source and should have (ProjectionOffsetX+
> SourceOffsetX,ProjectionOffsetY+SourceOffsetY,SourceToIsocenterDistance-SourceToDetectorDistance)
>
>
>
> Is this correct or did I miss something?
>
> Basically it comes down to the question: Do I need to compensate for the
> horizontal offset of the rotation axis in ProjectionOffsetX as well or is
> it already handled by setting SourceOffsetX?
>
>
>
>
>
> *Best,*
>
> *Moritz*
>
>
> _______________________________________________
> Rtk-users mailing list
> Rtk-users at public.kitware.com
> https://public.kitware.com/cgi-bin/mailman/listinfo/rtk-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/rtk-users/attachments/20220512/ec62e78d/attachment.htm>