Conflicts around the world leave mine-fields in need of clearing. The fields often hold buried metal debris in addition to mines, which complicates the task of detecting them. An excessive number of false alarms is time-consuming and may reduce operator care when dealing with a real mine. Cobham Technical Services wished to improve their signal processing techniques in order to reduce the number of false alarms, without increasing the risk of missing a mine.
Cobham Technical Services offered an industrial mathematics internship, under the scheme managed by the Knowledge Transfer Network (KTN) for Industrial Mathematics. A UEA PhD student, John Schofield, took up the sixth-month internship.
The work he undertook involved research into the problems of landmine detection, to try and improve the target detection rates of their mine detector. Landmine detection has the problem that all images suffer from the effects of diffraction caused by the wide beam width of the antennas used.
After considering various techniques, a multiple migration and scattering algorithm was proposed. This created various images, each with slightly different scattering profiles which were then overlaid in order to create a final image which only showed the target's true spatial location and no diffraction effects.
Further details about the project can be found in the KTN Project Case Study (PDF).
Some of the results from the research have since been published in the journal IEEE Transactions on Geoscience and Remote Sensing, as "A Multiple Migration and Stacking Algorithm Designed for Land Mine Detection".