Abstract

A variety of challenging tomography problems have emerged in recent years and their solution call for joined-up work on the underlying mathematics, the algorithms as well as the numerical software. With the hope to facilitate this process, we have developed the Core Imaging Library (CIL) – a python package for the solution of rich and “non-rich” tomography and other inverse problems (https://ccpi.ac.uk/cil/). As an example of rich tomography, I will describe our recent work on hyperspectral neutron tomography. Here, we developed a spatio-spectral reconstruction method in CIL for separating materials based on Bragg edges in energy-resolved neutron data. As an example of “non-rich” tomography, I will describe a directional total variation reconstruction method implemented in CIL and submitted as an entry to the recent Helsinki Tomography Challenge 2022 for limited-angle X-ray CT reconstruction. Finally, participants are invited to our upcoming training and hackathon to try out CIL for their own rich (and non-rich) tomography problems in Cambridge this March (https://ccpi.ac.uk/events/byod-cil-hackathon/).