The Key Techniques Research Of Forward-Looking Ground Penetrating

Forward-looking ground penetrating radar (FLGPR) is an effective tool to detect buried targets by discerning the discontinuities in the electric permittivity of the propagation medium. And it is a new type GPR developed on the base of down-looking GPR (DLGPR). FLGPR places its antennas in front of a mobile platform and inspects the ground surface of interest with an incidence angle.Compared with downward-looking systems, there are many advantages of FLGPR. FLGPR does not suffer from the problem of strong specular reflection from the ground, and is capable of collecting data for a much larger area in a much shorter time than DLGPR and imaging away from a safe distance. Furthermore, FLGPR can usually provide multiple observations on the same spot as the system moves forward, and can take advantage of multi-look processing to improve its detection capability. Due to these merits, the research of FLGPR has become an hot topic in the landmine detection community.Currently, the researches of FLGPR can be cast into the system hardware design and the signal processing technology. This dissertation focuses on the signal processing technology of FLGPR, which is a challenging task due to the following reasons: The landmine radar cross section RCS of FLGPR is smaller than that of DLGPR, which makes landmine detection by FLGPR a challenging problem. EM waves scatter from a random rough ground surface in unpredictable ways, contributing to clutter which distort and obscure the desired scattering field from a buried target. Problems such as refraction at the air/ground interface will also result in errors to simple propagation models because of the complexity in the propagation medium. The clutter reduction, synthetic aperture imaging and the electromagnetic wave velocity estimation are investigated in this dissertation. The contents and innovations of this dissertation are as following:1. Aimed at the clutter in radar data caused by backscattered from the scatterer external the imaging (survey) area in FLGPR, this dissertation proposes a method to reduce this kind of clutter. This clutter reduction method uses apex shifted parabolic Radon transform estimate the clutter parameter in the Radon domain. And the clutter are reconstructed by nonlinear curve fitting. Then the clutter is removed by subtraction method. By comparing the effect between pre-and post clutter reduction, the clutter reduction method increases the signal-to-noise ratio (SNR) of the radar images and improves the detection performance. And the clutter reduction method can obtain higher quality images.2. Two new methods, which are synthetic aperture imaging based on holographic imaging and synthetic aperture imaging based on nonstationary filter have been proposed. For the holographic imaging method, the propagation of electromagnetic wave in two different media can be equivalent to the propagation in one media by using the equivalent wave velocity, and the synthetic aperture imaging can be accomplished by holographic imaging method. But the nonstationary filtering method, by using nonstationary convolution filter, the spectrum of the refocused image can be reconstructed directly from the spectrum of backscattered signal from the target area. Furthermore, the theory of nonstationary filtering can be applied to the 3-D synthetic aperture imaging for FLGPR. The refraction correction of the target imaging can reduce the computational cost. The multi-look processing which improve its SNR and detection capability are accomplished by estimating the incidence angle of target.3. After analyzing the effects of the imprecise of wave velocity on the target locating. A method is proposed to estimate the velocity of the electromagnetic wave propagation under ground. This method estimates the velocity using the image gray variance after synthetic aperture imaging with various velocities. The effectiveness of the approach is demonstrated with simulated data and experimental data set.