search

CERN / TIGRE

TIGRE: Tomographic Iterative GPU-based Reconstruction Toolbox
BSD 3-Clause "New" or "Revised" License
581 stars 190 forks source link

Differences in reconstructed images #601

Open rasad-hub opened 4 weeks ago

rasad-hub commented 4 weeks ago

Hello. I tried to reconstruct the sinogram matrix I got from resolusion-phantom through tigre algorithms and I am sure that I enter the settings of the parameters correctly, but the reconstructed image is different from the image I expect. The photo on the left is related to the same sinogram. And it is reconstructed by a simple command (iradon) and the image on the right is reconstructed by tigre algorithms. Can you guide me, where is the problem?

AnderBiguri commented 4 weeks ago

The only guess I can give you is that you have set the geometry wrong somehow. It also seems that you have not log-transformed your projections either (on both cases).

AnderBiguri commented 4 weeks ago

yes, as I don't know if the angles match the measured data, or the geometry is right, it should match what the real thing is, which I don't know. Maybe try negating the angles, or other things.

rasad-hub commented 4 weeks ago

yes, as I don't know if the angles match the measured data, or the geometry is right, it should match what the real thing is, which I don't know. Maybe try negating the angles, or other things.

It is possible that this result is due to the tiger version or some mex files are not included during compilation?

AnderBiguri commented 3 weeks ago

@rasad-hub no, not at all. Both of your recons look wrong anyway, probably the log-transform, as I said. But TIGRE should work better than matlabs iradon.