Valeri Mikhnev

Single-reference near-field calibration procedure for step-frequency ground penetrating radar

The antenna system of ground penetrating radar is usually positioned in the vicinity of the ground surface to obtain sufficient spatial resolution for detecting small targets. Therefore, received signals are subject to errors caused by unwanted reflections and coupling. A near-field calibration procedure taking into account these effects as well as antenna dispersion is presented. For practical convenience, only a metal plane is used here as a calibration standard. Calibration data are measured for a set of distances between the antenna and metal plane. The reference reflections in all positions are computed using polynomial representation of the reflected signal. Then, error coefficients are obtained by a solution of a linear system of equations in the least squares sense. Experimental verification of the proposed calibration technique demonstrated efficient removal of artefacts caused by unwanted reflections. Moreover, peaks in the radar range profile become sharper after calibration if the antenna is dispersive. Consequently, the ground response in the form of an exponential term can be effectively subtracted from the received signal. This helps to reveal and discriminate shallow underground targets obscured by strong reflection of the ground surface. Experimental examples are given to illustrate the performance of the method.

Discrimination of Buried Objects in Impulse GPR Using Phase Retrieval Technique

This paper presents a modified signal processing technique of differentiating buried targets using impulse ground-penetrating radar. The technique is applicable to radargram; however, signal processing is carried out for each trace of the radargram separately. Every significant peak of each trace is compared with the peak of some calibration signal, for example, a signal that is reflected from the far interface of a thick rock slab. However, the calibration does not require the reference material be embedded in the medium, it can be in the air. The difference of phases between the object signal and the calibration signal is a function slowly varying with time that can be used for characterization of the buried object. The performance of the method is demonstrated in several examples using signals collected with the commercial impulse radar.

Subsurface target identification using phase profiling of impulse GPR data

The problem of target discrimination in ground penetrating radar data (GPR) is addressed in this study. Formerly developed frequency-domain technique based on building separate amplitude and phase profiles for every A-scan has been modified to be suitable also for impulse GPR data. To this end, impulse radar data have been preprocessed and transformed to frequency domain. The rest of the algorithm is rather similar to the case of frequency-domain radar data. The method yields a phase profile as a function of depth that is related to the phase shift occurring in the act of reflection of the wave from inhomogeneity existing at given depth. This phase shift depends on the contrast between the target and surrounding medium and thus can be used for target characterization. The technique was used to interpret data collected over a test site, where metal bars and a plastic pipe were buried inside sand and gravel. The radar data was collected with Malâs CX11 Concrete Imaging System operating at the 1600 MHz central frequency. The images built with the use of the novel signal processing technique demonstrate its validity for recognition of some buried objects.

A Fast 2-D Microwave Imaging Technique Using Synthetic Aperture Focusing of 1-D Profiles

A new microwave imaging technique forming the 2-D image from a set of one-dimensional dielectric profiles is proposed. To obtain the 2-D subsurface image, the backscatter data is collected along a line over the tested medium and processed by a conventional synthetic imaging method as done in the synthetic aperture radar (SAR). In this work, every radar range profile is first transformed to one-dimensional dielectric permittivity distribution. Then the obtained dielectric profiles are combined using SAR to build the 2-D image. This technique though cannot be classified as a true two-dimensional inverse scattering method, yields quite reasonable image being much better than using SAR processing alone. Besides, this technique is rather fast and thus suitable in practical subsurface radar systems. As reconstruction of the one-dimensional permittivity profile takes about one second of a standard personal computer time, the whole 2-D image can be obtained in several seconds

A modified bow-tie antenna design for step-frequency subsurface radar applications

A modified bow-tie antenna for step-frequency ground penetrating radar is proposed. The flared dipole arm consists of a metal sector gradually converting into a strip-slot structure loaded by resistive films. For the comparison, a TEM horn is used. The synthesized pulse waveform is computed from the frequency-domain data using a calibration procedure and inverse Fourier transform. The antenna performance is demonstrated in several examples of detecting small shallowly buried subsurface targets.