Multi algorithm based landmine detection using ground penetration radar

The principle of Ground Penetrating Radar (GPR) is based on the scattering of electromagnetic waves at subsurface with reflection and refraction from the buried object due to change in material. This paper deals with analysing the GPR data, to detect anti-tank landmines buried in various conditions. This requires reading GPR data and plottingxy, yz, and zx plane and then normalizing the data followed by implementing algorithms for background removal processing &noise reduction. First normalizing a GPR data, to minimize the influences due to clutter effect, deployment method and experimental setup. Next, for the noise reduction, mean subtraction method & median filter is developed, Gaussian filter developed which helps in estimating the number of objects and image smoothing function used to reduce the noise. In this paper, an approach is proposed for detecting landmines buried in sand using a GPR.