Data for paper: Adaptive filtering method for reducing signal saturation induced artifacts in X-ray dark-field tomography

This dataset includes raw object images, raw reference images, extracted projections, and reconstructed volumes of a mini-plum tomato sample (imaged in a cylindrical plastic tube filled with water). The dataset was created for the purpose of demonstrating an adaptive filtering method for reducing saturation-induced artifacts that appear in X-ray dark-field and phase-contrast data in the presence of highly-scattering structures. Link to published paper: TBA.

- The raw data and scan parameters/info files ("params.txt", "scan.csv") are found in the zipped folder "raw.zip".
- The folder "raw_angle_150" includes raw images (phase stepping sets) and scan parameter files for the tomographic angle 150 (also included in the zip file). This can be used for quick testing or viewing.
- The raw phase stepping images (.smv) are binary images of type uint16 (little-endian) with a header size of 512, width of 1536 and height of 864 pixels.
- Processed/extracted transmission (T), phase-contrast (P) and dark-field (D) projections are found under the folder "processed" for different minimum SNR values used in the adaptive filtering (0 is unfiltered). Each projection angle has its own .h5 file. Each h5 file has datasets (864x864, float64) for each of the image signals (+ corresponding SNR_D image, named "vis_map").
- Reconstructed volumes corresponding to each minimum SNR value are found under the "reconstructed" folder, along with corresponding configuration files used in the reconstruction process. The raw volumes of type 32-bit real (little endian) have a size of 512x512x624.