[Rtk-users] artifacts on reconstruction

[Simon Rit]

Dear Julia,
Interesting, is this a crab? I agree with you, this is a geometry problem.
My guess is that the projections are not adequately centered. If you use
rtksimulatedgeometry to produce the geometry file for RTK, I would play
with --proj_iso_x to correctly set the position of the projection of the
rotation axis in the projections. Hopefully, this information is contained
in your geometry information.
I'm not surprised that --sid has mainly a magnification effect. I don't
think sid/sdd is the main problem here.
We can try to help if you can't figure it out but you'd have to send the
data.
Simon

On Tue, Mar 15, 2016 at 7:19 AM, Julia Semyakishkina <
prekrasnaya1985 at gmail.com> wrote:

> Hello. I use RTK and another tomography packets/libraries for
> reconstruction and comparison results. RTK is one of the best for me, but
> with one model/sample i have strange artifacts, which doesn't appear in
> another packets.
> Here http://i.imgur.com/U9hqc6v.png is one projection. Here
> http://i.imgur.com/MOlygHS.png is one slice reconstructed by RTK. Here
> http://i.imgur.com/4a3jtLm.png is the same slice reconstructed by another
> packet.
> I am sure that i set geometry information correctly. Another models with
> the same tomograph, same geometry, same acquisition settings reconstuct
> fine and even better then in another packets. But exact reconstruction of
> the bug's foots caused to the artifacts.
> Even so, i think that is geometry problem. But i don't know how to solve
> it in RTK. I played with sid parameter in another packet and got very
> familiar artifacts, which appear with correct geometry in RTK. Just for
> test i played in the same way with RTK, i.e. change sid to wrong values and
> nothing changes except scale. Im a bit confused of this fact, i dont
> understand how sid\sdd doesn't make a sense in reconstruction, indeed when
> sid changes, beam angles which go through the model change, and from here
> projection shall be different.
> I can upload raw data with all meta info if needed.
> Any advice or direction where i should go to solve the problem would be
> appreciated.
>
> _______________________________________________
> 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/20160315/dccba72e/attachment-0010.html>