Prediction of GPR performance in landmine detection

Abstract

The contrast in the dielectric constant between a landmine and the surrounding soil is one of the most important parameters to be considered when using ground penetrating radar (GPR) for landmine detection. In this paper, we discuss available models for the prediction of the dielectric constant from soil physical properties including bulk density, particles density soil texture, and water content. We predict the effects of such properties on the antipersonnel (AP) landmine detection performance of GPR in an application in Iran. Initially, available soil geophysical information was used from four types of soil selected from Iranian mine-affected areas. Subsequently, a pedotransfer model was developed to predict whether or not field conditions are appropriate for use of GPR instruments. The predictions outcome obtained through usage of this model was based on different soil textures at various soil water contents. Knowledge of soil texture, dry bulk density, and water content are necessary to determine whether soil conditions are suitable for utilization of GPR mine detection. The developed model presented here can be useful for making this determination. Finally, the graphical user interface (GUI) of the pedotransfer model was calculated and presented herein. This software package facilitates the analysis of complex dielectric constant of soil as well as attenuation of GPR signals. The developed package is also capable of plotting the complex dielectric constant of soil coupled with attenuation of GPR signals versus soil physical properties.