The issue with slice image display

[daxiaojiupang]

Figure 1 shows the slice image from the FBP algorithm in d15_3Dparaller.m, while Figure 2 displays the slice image from the OS_SART algorithm in d15. The projection data is my own. The FBP-reconstructed slice clearly shows the object, but why is the iterative algorithm's reconstructed image entirely black with nothing visible?

[AnderBiguri]

It does look like you have not applied the beer-law (log transform) to your projection data.

It also looks like you may have a static machine (probably a syncrotron of some sorts) and a rotating stage. This will need some geo.COR corrections, to correct for the offset of the axis from the center point.

[daxiaojiupang]

Thank you very much for your reply. I directly used the projection data obtained from the CCD camera for reconstruction. My setup features a fixed radiation source and a CCD camera, with the object placed on a rotating stage. Based on the issues you pointed out, how should I modify my approach? Thank you.

…

---- Replied Message ---- | From | ***@***.***> | | Date | 06/03/2025 16:47 | | To | ***@***.***> | | Cc | ***@***.***>***@***.***> | | Subject | Re: [CERN/TIGRE] The issue with slice image display (Issue #661) | AnderBiguri left a comment (CERN/TIGRE#661) It does look like you have not applied the beer-law (log transform) to your projection data. It also looks like you may have a static machine (probably a syncrotron of some sorts) and a rotating stage. This will need some geo.COR corrections, to correct for the offset of the axis from the center point. — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: ***@***.***>

[AnderBiguri]

@daxiaojiupang check the demo for real data, but in essence, you capture X-ray transmission. For recosntruction, you need X-ray absorbtion.

p=-log(p/max_val) where max val is the maximum possible pixel value, often 2^16.