The feasibility study of the landmine identification system (LIS) based on the angular distribution of thermal neutrons for buried landmines was undertaken in this work. The operation of the system was based on both artificial neural networks (ANN) and least-squares methods in which the input data were prepared by the MCNP6.1 Monte Carlo code before feeding into MATLAB and dedicated programs, respectively. Having achieved the promising ANN and least-squares results (with a relative error of less than 15%), the study confirmed that the proposed system is sensitive to the depth of the landmine as well as to the soil moisture, therefore it could be used for landmine identification.

Keywords