表层穿透雷达精细成像技术研究

发布时间：2018-04-11 21:15

本文选题：表层穿透雷达 + 精细成像　；参考：《国防科学技术大学》2014年硕士论文

【摘要】：表层穿透雷达以其穿透介质实施观测的优良性能,正日益为人们所重视,在安检防暴、扫雷探测、无损评估(Nondestructive Evaluation,NDE)等诸多军事和民用场合得到了广泛应用。表层穿透雷达成像技术作为直观呈现探测结果、降低数据解译难度的重要手段,发展却相对滞后,离实用化的要求还有一定差距。本文提出了表层穿透雷达精细成像的概念,核心是高分辨成像、改善适应性以及增强稳健性,以期提高表层穿透雷达的成像性能,推动其实用化发展。本文研究表层穿透雷达精细成像技术,主要包括:表层穿透雷达成像基本原理、快速自聚焦方法和高分辨成像方法。首先,介绍表层穿透雷达成像机理。表层穿透雷达成像属于近场成像问题,因此基于平面波近似的成像算法并不适用,必须在充分考虑散射场的波动性的基础上研究成像算法。第二章利用波动方程,对波恩近似下的散射场表示形式进行分析,研究近场条件下的三维距离偏移算法和全息成像算法,并对算法性能和空间采样准则进行讨论。其次,研究表层穿透雷达快速自聚焦成像方法。在工程应用中,成像场景的参数常常难以准确获取,成像结果容易受参数误差影响而发生模糊、分辨率下降等问题。为增强成像算法对环境条件的适应性,减轻对人工干预的依赖,第三章研究了自动调整和修正成像参数以实现准确聚焦成像的方法。本文首先介绍基于幅度和的自聚焦算法,通过合理地选择聚焦度量来减少运算量,在最优化聚焦度量过程中搜索最优成像参数。然后,从信号的卷积退化模型来重新认识散焦问题,提出一种基于反卷积去除参数误差的影响来实现自聚焦的算法。最后,研究表层穿透雷达高分辨率成像技术。表层穿透雷达方位向分辨率通常受到合成孔径范围、工作频率等的制约,可以通过拓宽频谱支撑区的方法来提高成像分辨率。第四章研究了两种拓展频谱支撑区来提高成像分辨率的方法。一种是基于自回归模型的谱外推方法,利用线性预测模型来计算支撑区外的频谱成分,从而获取额外的带宽。另一种是反卷积校频谱均衡的高分辨成像方法,通过消除天线方向图对高频分量的抑制作用,均衡高低频成分,达到扩展支撑区宽度的目的。本文通过构建的实验系统收集实测数据验证了所研究的各种算法,实验结果显示这些算法对于提高成像质量起到了重要作用。本文所研究的表层穿透雷达精细成像技术,为提高表层穿透雷达成像性能、增强系统的适应性和稳健性打下了基础,推动了表层穿透雷达的实用化进程。
[Abstract]:Surface penetrating radar has been widely used in many military and civil fields, such as riot control, mine clearance detection, nondestructive evaluation, and so on, because of its excellent performance of observation in penetrating medium.Surface penetrating radar imaging technology as an important means to visualize the detection results and reduce the difficulty of data interpretation, the development of surface penetrating radar imaging technology is relatively lagging behind, and there is still a certain gap from the practical requirements.In this paper, the concept of fine imaging of surface penetrating radar is proposed, the core of which is high resolution imaging, improving adaptability and enhancing robustness, in order to improve the imaging performance of surface penetrating radar and promote its practical development.In this paper, the fine imaging technology of surface penetrating radar is studied, including the basic principle of surface penetrating radar imaging, fast self-focusing method and high-resolution imaging method.Firstly, the imaging mechanism of surface penetrating radar is introduced.Surface penetrating radar imaging is a near-field imaging problem, so the imaging algorithm based on plane wave approximation is not applicable. Therefore, it is necessary to study the imaging algorithm on the basis of fully considering the fluctuation of scattering field.In the second chapter, the representation of scattering field in the Bonn approximation is analyzed by using the wave equation, and the 3-D distance migration algorithm and holographic imaging algorithm under near-field condition are studied, and the performance of the algorithm and the spatial sampling criterion are discussed.Secondly, the fast self-focusing imaging method of surface penetrating radar is studied.In the engineering application, the parameters of the imaging scene are often difficult to obtain accurately, the imaging results are easily affected by the parameter error and the resolution is decreased.In order to enhance the adaptability of the imaging algorithm to the environmental conditions and reduce the dependence on human intervention, chapter 3 studies the method of automatically adjusting and modifying the imaging parameters to achieve accurate focus imaging.In this paper, we first introduce a self-focusing algorithm based on amplitude sum. By selecting the focusing metric reasonably, we can reduce the computational complexity and search for the optimal imaging parameters in the process of optimization focusing measurement.Then, the defocusing problem is re-recognized from the signal convolution degradation model, and a self-focusing algorithm based on the effect of deconvolution removal parameter error is proposed.Finally, the high resolution imaging technology of surface penetrating radar is studied.The azimuth resolution of surface penetrating radar is usually restricted by the range of synthetic aperture and the working frequency.In chapter 4, two methods to improve imaging resolution are studied.One is the spectral extrapolation method based on the autoregressive model, which uses the linear prediction model to calculate the spectrum components outside the support area, thus obtaining the extra bandwidth.The other is the high resolution imaging method of deconvolution correction spectrum equalization. By eliminating the suppression effect of antenna pattern on the high frequency component, the high and low frequency components can be equalized to achieve the purpose of extending the width of the support region.The experimental results show that these algorithms play an important role in improving the imaging quality.The fine imaging technology of surface penetrating radar studied in this paper has laid a foundation for improving the imaging performance of surface penetrating radar, enhancing the adaptability and robustness of the system, and promoting the practical process of surface penetrating radar.
【学位授予单位】：国防科学技术大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN957.52

【参考文献】