Hi Simon,

thank you very much for your quick reply.  Your analysis was correct, I needed to redefine the center.

I wish you a very nice day,

best regards,

Vincent



On 2018-10-01 20:43, Simon Rit wrote:
Hi Vincent,
This generally occurs when your detector offset is not adequately passed to the geometry. When you say that you have "5.2mm and -8mm" offsets, I guess it is with respect to a centered geometry, where the central ray hits the detector. For this to be the case in RTK, you must check that the coordinate (0,0) mm corresponds to the center of the detector. By default, I guess that tif puts it in the corner. So you probably need to add to rtkprojections an option like --neworigin xorig,yorig,0 where xorig and yorig is something like -0.5*(npixels-1)*spacing. If that doesn't help, you don't need to share the data, the mhd headers are sufficient to test the geometry.
Good luck!
Simon

On Mon, Oct 1, 2018 at 5:13 PM vincent <[email protected] <mailto:[email protected]>> wrote:

    Hi everyone,

    I am facing a problem when I try to reconstruct an object. I'll
    try to
    be as complete as possible in the description.

    I acquired 18000 views from a custom-made ct scanner, of which I
    extracted 360 for the sake of reconstruction speed for my tests. The
    images are cropped to 1001*1761 pixels.  An external calibration
    program
    estimated that the x and y detector offsets where 5.2mm and -8mm
    respectively.  The sdd and sid are 983mm and 483mm.  I used those
    value
    to generate a simulated geometry.

    I then created a projection volume with the command

    rtkprojections -v  -o rec1Proj.mha -p . -r cropped.*.tif --i0 7000
    where
    I estimated the i0 value on the images background (for the sake of
    completeness, I also tested with the automatic value which
    resulted in
    the same problem)

    Finally, I wanted to reconstruct the volume with the command

    rtkfdk  -g geoCT.xml -p . -r rec1Proj.mha  --spacing 0.2 --dimension
    1761,1761,1761 -v -o recFDK.mha  --hardware cuda


    The exact error message is:

    Reconstructing and writing... ExceptionObject caught with
    writer->Update() in file
    ~/IPTools/RTK/RTK/applications/rtkfdk/rtkfdk.cxx line 218

    itk::ExceptionObject (0x1cf4b90)
    Location: "unknown"
    File: ~/IPTools/RTK/RTK/code/rtkDisplacedDetectorImageFilter.hxx
    Line: 155
    Description: itk::ERROR: Cannot account for detector displacement
    larger
    than 50% of panel size. Corner inf=2.55504 and corner sup=100.826

    I had a look at the aforementionned line in the
    rtkDisplacedDetectorImageFilter.hxx file but I didn't really
    understand
    how the corners are computed in the
    
inputPtr->TransformIndexToPhysicalPoint(inputPtr->GetLargestPossibleRegion().GetIndex(),

    corner function.

    I have no problem sharing the projections if needed, but they
    weight 1.3
    Go altogether.


    I thank you very much in advance for any kind of help you can provide.

    Best regards,

    Vincent


    _______________________________________________
    Rtk-users mailing list
    [email protected] <mailto:[email protected]>
    https://public.kitware.com/mailman/listinfo/rtk-users


_______________________________________________
Rtk-users mailing list
[email protected]
https://public.kitware.com/mailman/listinfo/rtk-users

Reply via email to