Research Of Fine Surface Penetrating Radar Imaging

Being capable of penetrating medium and detecting subsurface targets, the Surface Penetrating Radar(SPR) has drawn more and more attention and is applied in a wide spectrum of military and civil use including security checking, minesweeping, and Nondestructive Evaluation(NDE). However, the Surface Penetrating Radar Imaging technique, an important approach to present and interpret the detecting result, is still inadequately studied and unpractical. To solve this problem and improving the imaging quality of SPR, this thesis developed the method of Fine SPR Imaging aiming at improving resolution, applicability as well as robustness.The main content of this thesis is composed of three parts: the basis of SPR imaging, fast autofocus method and high resolution imaging method.First, the principle of SPR imaging was presented in Chapter 2. Due to the nearfield conditions, the imaging algorithm base on the plane wavefront was no longer feasible. The wave properties of scattered field must be considered carefully. Thus, the wave equation was adopted to analyze the scattered field and to derive the three-dimension Rang Migration(RM) algorithm.Second, the fast autofocus algorithms were studied. In practical terms, the accurate parameter of imaging area is hard to acquire and the inaccuracies of parameters lead to the degradation of imaging result. To enhance the applicability to different imaging scenarios and to reduce the dependence on experts, this thesis developed autofocus algorithms to adjust and rectify imaging parameter automatically. In Chapter 3, the mathematic model of autofocusing was studied and effect of the unknown parameter on imaging result was revealed. Two autofocus algorithms were presented. One was based on the focusing metrics and searched the correct parameter by optimizing the focusing metrics. The computational efficiency was specially optimized. The other algorithm regarded the out-of-focus as the degradation of image caused by convolution and promoted a deconvolution algorithm to eliminate the effect of parameter estimation error and then to focus the image.Third, the high resolution imaging algorithms were researched. The lateral resolution of SPR is tend to be constrained by factors such as the length of synthetic aperture and the operating frequency. Chapter 4 presented two algorithms to improve resolution. The bandwidth was extrapolated by AR model to enhance the resolution in the first algorithm. The second algorithm took the effect of antenna into account. Using the antenna pattern to deconvole the recorded data and the suppression of high frequency component caused by antenna pattern was removed.Practical measurement data was collected with an experimental imaging system to verify the capability as well as feasibility of the presented algorithms. The importance of the precise SPR imaging method studied in this thesis lies in improving resolution, enhancing the applicability and making the SPR imaging technique more practical.