Computed tomography image processing and reconstruction using the ASTRA toolbox

Thumbnail Image

Date

2022

Authors

Chinamatira, Gideon

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

The computed tomography (CT) technique has become pivotal to research in fields such as palaeontology, health studies, engineering, and materials science. This imaging technique makes use of x-ray absorption in order to obtain three-dimensional (3D) images. However, imaging can be challenging when objects of high-density materials are included in the matrix with lower-density material. Extremely bright regions, streaks, and scattering can result in such cases thereby reducing contrast and also making the data less usable. There are several algorithms that are used at the stage of image reconstruction when the 3D volume is generated from radiographic projections such as the filtered back projection (FBP). Backprojection has two distinctive limitations, noise, and streak artefacts and it is a combination of these restrictions and the advancement of computers that iterative algorithms are seen as the possible candidate to replace this method of image reconstruction to subsequently reduce these artefacts. We report an investigation on the effects of adding an image processing step before performing computed tomography reconstructions. We have acquired projections with a micro-computed tomography scanner and carried out image processing functions for the improvement of the quality of the data output using a paleontological specimen with a significant amount of iron inclusions which cause bright and dark streak artefacts. We have applied several filters to alter the projections before reconstruction. The results so far show that the minimum filter, median and the mean filter reduces the noise and streak artefacts in the specimen. The Gaussian smoothing filter also successfully reduces the noise in the images, but the streak artefacts are still significantly visible. The unsharp mask filter enhances the edges in the images and reduces the streak artefacts significantly. However, this filter enhances other high-frequency components in an image causing an increase in the noise.

Description

A dissertation submitted in fulfilment of the requirements for the degree of Master of Science to the Faculty of Science, School of Physics, University of the Witwatersrand, Johannesburg, 2022

Keywords

Citation

Collections

Endorsement

Review

Supplemented By

Referenced By