[Rtk-users] Flat Field Image Correction

[Simon Rit]

Hello,
Usually, scanners do this for you and you don't have to do it manually, 
we only use constants for i0 (--i0) and dark (--idark). If you have to 
do it, you can use itk to do it projection per projection. The filters 
you use are
itk::SubractImageFilter (difference between two images)
itk::DivideImageFilter (ratio between two images)
itk::LogImageFilter
I've done this for some cases, e.g., XRadRawToAttenuationImageFilter, 
look at the code.
Simon

On 22/04/2016 13:27, AMJAD N wrote:
> Hello users,
>
> I have read Flat field image correction using dark ( projection with 
> CT-off, detector-on ) and bright ( projection with CT-on, detector-on 
> and no object in between CT and detector ) images procured before 
> scanning any substance.
>
> correction = log (bright image) - log ( projection-dark image ); // 
> for all projections
>
> This pre-processed images are then fed to FDK algorithm.
>
> Did someone here tried this method? If I want to implement this, do I 
> have to write a separate class for this? Any suggestions for that?
>
> PS: Presently I am using LUTbasedVariableI0RawToAttenuationImageFilter 
> and I am getting reasonably good output. When i implemented FDK in 
> MATLAB i used this flat field correction and result was good. So I 
> want to know the impact with RTK.
>
> Thanks and Regards,
>
> AMJAD N
>
>
> _______________________________________________
> Rtk-users mailing list
> Rtk-users at public.kitware.com
> http://public.kitware.com/mailman/listinfo/rtk-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/rtk-users/attachments/20160422/7e722350/attachment-0010.html>