Advanced imaging by space-time deconvolution in array GPR

Abstract

Digital beamforming in array-based UWB radar delivers a high-resolution 3-D image of subsurface in GPR landmine detection while simultaneous data acquisition by elements of the array significantly increases the scanning speed. Such a GPR system with a single transmit antenna and a linear receive array has been developed in the Delft University of Technology. For online processing we propose an advanced imaging algorithm based on migration by regularized, parametric space-time deconvolution. The algorithm deconvolves a 3-D space-time array point spread function out of the data volume by means of FFT and inverse Wiener filter that is being controlled automatically with numerical criteria for stability and accuracy. The prior knowledge of GPR impulse response and ground impulse response is used to form a separate point spread function for each receiving antenna. The developed technique has been verified on experimental data for typical anti-personnel mines and compared with a classical migration by diffraction stacking. © Springer Science+Business Media, LLC 2010.