Inversion Of Geometrical Properties Of Underground Targets Using Polarimetric Ground Penetrating Radar

Ground penetrating radar(GPR),as a shallow surface subsurface target detection method,has the advantages of lossless,high resolution and fast acquisition.It has been widely used in detecting underground pipelines,concrete structures,voids,etc.At present,the research is lack in the data processing of ground penetrating radar data so that a large amount of information in radar data is wasted.In this thesis,a series of studies on data detection and data processing of polarized GPR systems are carried out:This thesis proposes a hybrid polarimetric GPR system which can obtain more polarization information than traditional commercial ground penetrating radar.The hybrid polarimetric GPR system designed in this thesis consists of a vector network analyzer,a PC control unit and a hybrid polarized antenna array.The antenna array of the hybrid polarimetric GPR system consists of a right-handed planar helical antenna for transmission and two orthogonal Vivaldi antennas for receiving.The right-handed planar helical antenna emits a right-handed circularly polarized wave and the two orthogonal Vivaldi antennas receive mutually perpendicular linearly polarized waves to satisfy the mixed polarization mode.In this thesis,a polarization calibration method for hybrid polarized antenna arrays is proposed.A gridded trihedral is used in the polarimetric calibration experiment to correct the imbalances and cross-talk between the transmitting channel and the two receiving channels.The method is effective according to the result of this experiment.In terms of imaging algorithm,this thesis performs migration imaging on radar data.Migration imaging can correct the position of targets,reconstruct the images and improve the lateral resolution.Back projection imaging algorithm is used in this thesis.This thesis proposes a method for polarimetric decomposition,namely the method of full polarization target identification based on entropy(that is,H-alpha polari'zation decomposition method).This method is suitable in multi-category target classification which can divide targets into eight categories.An experiment of a metal sphere,a scatterer and a dihedral and a tube is carried out in our thesis.The result of this experiment verified the feasibility of H-alpha polarization decomposition method in GPR applications.In the azimuth estimation of the elongated obj ect,this thesis adopts a method based on Alford Rotation for azimuth estimation.The scattering matrix can be extracted from the received signals.Based on the scattering matrix of the elongated object,the azimuth can be estimated by the mathematical rotation of the measurement frame.A laboratory experiment was further conducted to detect five metal rebar buried in dry sand at different orientation angle relative to the GPR scan direction.Based on the result of this experiment.It is concluded that radar polarimetry can provide not only richer information than single-polarization GPR,but also a reliable approach for orientation estimation of a subsurface elongated object.The measurement range is 0°-180°.