Hyperspectral X-ray CT datasets of an aluminium phantom containing three metal-based powders (original) (raw)
2020
Abstract
<strong>General Data description:</strong> This is a set of two hyperspectral (energy-resolved) X-ray CT projection datasets of a multi-phase phantom. It was acquired in a custom-built, laboratory micro-CT scanner with an energy-sensitive HEXITEC detector in the Henry Moseley X-ray Imaging Facility at The University of Manchester. The following data contains all the files necessary for reconstruction, following two hyperspectral scans of a metal, multi-phase phantom. The phantom consists of an external aluminium cylinder, with three holes, each filled with a different metal-based powder (CeO<sub>2</sub>, ZnO, Fe). Each powder provides a unique attenuation signal, with CeO<sub>2</sub> in particular producing a distinct spectral marker which can be measured by an energy-sensitive detector. Two identical scans were acquired, with only the exposure time per projection changed. <strong>File descriptions:</strong> Contained is an image (.jpg) of the sample, along with five MATLAB (.mat) data files, as well as a single text (.txt) file. Where necessary, the files have been named to match the dataset they belong to, based on the different exposure times used for each dataset. Phantom_design_measurements.jpg contains a photograph of the physical phantom, combined with a diagram showing full sample measurements. Powder_phantom_scan_geometry.txt gives a breakdown of the full sample and detector geometry used when acquiring the raw projections for both scans. powder_phantom_30s_sinogram_full.mat contains the full 4D sinogram constructed following flatfield normalisation of the raw projection data, where an exposure time of 30 s was used for each projection. The 4D array contains the total number of energy channels acquired during scanning, followed by vertical and horizontal pixel number, and finally total projections angles acquired during scanning. powder_phantom_180s_sinogram_full.mat is the full 4D sinogram for the dataset, when exposure times of 180 s were used for each projection, following flatfield normalisation. A d [...]
Robert Cernik hasn't uploaded this paper.
Let Robert know you want this paper to be uploaded.
Ask for this paper to be uploaded.