CT reconstruction and image processing pipelines for cultural heritage Synchrotron X-ray Computed Tomography (SXCT) scans at beamline ID10-BEATS of SESAME. All pipelines can be found in the Image processing notebooks folder.
- By Gianluca Iori, Philipp Hans, 2024
- Code licence: MIT
- Narrative licence: CC-BY
- Created on: 05.05.2024
- Last update: 03.01.2025
| Notebook | Description | Binder URL |
|---|---|---|
| BEATS_recon_Roman_glass.ipynb | Phase-contrast SXCT reconstruction pipeline |  |
| BEATS_image_filtering_pipeline.ipynb | Image filtering facilitating segmentation of large 3D volume | |
| BEATS_egyptian_blue_pore_size.ipynb | Plot and visualize results from analysis of porous material | |
| Beamline information | |
|---|---|
| Beamline | ID10-BEATS@SESAME |
| Beamtime | In-House research |
- The underlying data for this submission is proprietary and subject to embargo according to SESAME's data policy.
- The pipeline for reconstruction of phase-contrast SXCT images can be reproduced using the following open-source dataset from SESAME ID10-BEATS.
| Dataset name | Description | DOI |
|---|---|---|
bee_yazeed-20231001T170032.h5 |
SXCT scan of a wasp |  |
| Sample | Egyptian blue |
|---|---|
| Scan name | egyptian_blue-20240229T135258 |
| Energy | 45 keV |
| Detector | Det 2 (Hasselblad system) |
| Camera | PCO.edge 5.5 |
| Voxel size | 3.1 um |
| SDD | 300 mm |
| Sample | Roman glass |
|---|---|
| Scan name | glass_room-M_stitch-20240222T153555 |
| Energy | 20 keV |
| Detector | Det 2 (Hasselblad system) |
| Camera | ORYX FLIR 7.1 MP GigE |
| Voxel size | 4.5 um |
| SDD | 250 mm |
| Field of view extension | 360-degree x 3 stitch scans |
The reconstruction pipeline relies on TomoPy and ASTRA. Instructions on how to set up a reconstruction conda environment for TomoPy can be found here and here.
Tip
At SESAME BEATS, we installed and built ASTRA and TomoPy from source in a dedicated conda environment. Instructions and a list of dependencies can be found here.
A minimal list of dependencies for 3D image processing in Python can be found here.