Hand-Held GPR Imaging Using Migration for Irregular Data

Xuan Feng,Motoyuki Sato,Cai Liu
DOI: https://doi.org/10.1109/jstars.2011.2164573
IF: 4.715
2011-01-01
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Abstract:Handheld ground-penetrating radar (GPR) system is one of a number of technologies that has been researched as a means of improving landmine detection efficiency. But it is difficult to display landmine imaging in horizontal slices, because human being operator cannot collect regular GPR data along the survey line and the measurement positions are random. Also detection of buried landmines by GPR normally suffers from very strong clutter that will decrease the image quality. So we proposed two processing procedures to achieve hand-held GPR subsurface imaging. One processing procedure is composed of interpolation and migration algorithm. The interpolation is chosen to produce the regular grid data set. The migration is used to improve signal-clutter ratio and reconstruct subsurface imaging. In the other processing procedure we developed a simply modified migration algorithm that can directly processes irregular GPR data. Two processing procedures, migration with interpolation and modified migration without interpolation, were applied to a field experiment data respectively. Both of processing procedures can efficiently process irregular GPR data and achieve subsurface imaging of high quality. The imaging processed by migration with interpolation is used in the dual sensor, advanced landmine imaging system (ALIS), and two sets of ALIS have detected more than 80 AP-mines since 2009. Also the modified migration algorithm can save the processing time, which is important for field real-time signal processing. We concluded that to achieve clear hand-held GPR imaging, the interpolation is not necessary algorithm and the migration is the key algorithm.
What problem does this paper attempt to address?