Our research in inverse problems and imaging

Our research in the area of inverse problems and imaging is devoted to improving existing methodologies and developing new approaches for the identification of hidden targets from electromagnetic field measurements.

The identification of hidden targets has many important applications including:

  • Improvements to airport and event security
  • Detection of unexploded ordnance
  • Improvement of food safety by the detection of foreign bodies
  • Improving medical imaging for more accurate patient diagnosis

Current EPSRC funded interdisciplinary collaborative work involves teams of engineers, computer scientists, mathematicians and medical practitioners.

Research is undertaken in partnership with Bath University, Cardiff University, University of Manchester and University of South Wales. Industrial partners including ABM University Health Trust.

These projects are aimed at finding new ways of classifying hidden targets and improving computational and algorithmic approaches for locating and characterising the objects.

Research areas

Find out more about our research in inverse problems and imaging:

Magnetic induction tomography

From September 2013 a three-year collaborative EPSRC funded project (EP/K023950/1), with Professors Brown, Marletta and Walker from the Schools of Mathematics and Computer Science & Informatics at Cardiff University, will investigate harnessing the power of GPUs and parallel processing to develop new approaches for the reconstruction of images from low frequency magnetic induction field measurements.

Here the inverse problem relates to determining the regions of higher conductivity in a lower conductive background from measurements of the perturbed field.

The project will involve theoretical, computational as well as engineering aspects applied to the relatively new medical imaging modality, magnetic induction tomography.  

This application could potentially offer advantages for early stages and monitoring of strokes, which are associated with contrasts in the electromagnetic parameters.

The project will work together with its collaborative partners at Bath University, University of South Wales and ABM University Health Trust, to ensure maximum impact of the planned developments.

For more information click here.

Polarization tensors

In July 2013 a two-year collaborative EPSRC funded project (EP/K039865/1) with Professor Lionheart, School of Mathematics, University of Manchester, started. The project intends to build links with the world leaders in the field of characterization of objects by means of asymptotic formulae and polarisation tensors. This mathematical description allows the shape and material properties of the inclusion to be described and offers possibilities for the low-cost solution of inverse problems.

Currently the focus is the metal detection problem and investigating how the sensitivity of field measurements due to magnetic induction associated with a conducting object can be expressed in terms of new classes of polarization tensors. The ability to form a library of these tensors for different objects has important implications for the identification of metallic targets in applications such as airport and event security, the detection of unexploded ordnance, improvement of food safety by the detection of foreign bodies and the location of archaeological artefacts.

Links have been established with Professor Ammari, Ecole Normale Supérieure, France; Professor Kang, Inha University, South Korea; Professor Peyton, University of Manchester; and Professor Volkov, Worcester Polytechnic Institute, USA.

Click here to find out more.

Finite element mesh of a simulated MIT system

Finite element mesh of a simulated MIT system

Includes human head and current and measurement coils, which contains 82,852 tetrahedral elements and was developed as part of the previous LCOMIT EPSRC grant EP/E009697/1.