近场高分辨三维SAR成像理论与方法研究

Near field high resolution three-dimensional (3D) SAR imaging becomes a research hotspot in recent years. This project aims to overcome the deficiencies that exist in the system structure and imaging algorithms of near field 3D SAR imaging. The imaging mechanism of near field 3D SAR is analyzed first. Combining the SAR imaging principle and electromagnetic measure theory together, the echo model of near field high resolution 3D imaging is built. Then using the electromagnetic spectral analysis method, the exponential term in the echo model which represents a spherical wave is decomposed into a superposition of plane wave components. Based on the decomposition, the mathematical model of near field 3D SAR is developed. To scan the target omnidirectionally, a system structure model that uses the circular synthetic aperture of single-antenna is adopted. At last, the high resolution 3D SAR image of near field target is obtained by using the spherical wave based Green function Fourier analysis method. The sub-aperture data segmentation technique and overlapped sub-aperture splicing technique are also proposed. The effectiveness of the proposed 3D imaging method will be verified by computer simulation. The research of this project will make contribution to the development of SAR imaging in various aspects, including the theory, system structure and imaging algorithms. It will also promote engineering applications of the near field 3D SAR imaging technique such as security inspection and anti-terrorism.

近场高分辨三维SAR成像是当前雷达成像领域的研究热点，针对近场三维SAR成像研究中系统结构和成像算法上存在的不足，本课题研究从近场三维SAR成像机理出发，结合SAR成像原理以及电磁波近场测量理论，建立近场高分辨三维SAR成像回波模型。运用电磁波谱理论分析方法，将近场条件下SAR回波模型中由球面波函数构成的指数项变换为平面波叠加的形式，继而建立近场三维SAR数学模型。提出了采用圆周合成孔径单天线的系统结构模型对目标进行全方位扫描，同时采用基于球面波的格林函数傅里叶分析方法来实现对近场目标的高分辨三维成像。基于数据分割和图像拼接理论提出了子孔径数据分割方法和重叠子孔径图像拼接技术，并通过计算机仿真，验证三维成像算法的有效性。开展本课题研究工作，有望在近场三维SAR成像理论、系统结构和成像算法方面获得创新性成果，并推动近场三维SAR成像技术在安检、反恐等领域的工程应用。

三维 SAR 成像技术是在传统二维 SAR 成像基础上发展起来的，该技术不仅具有传统 SAR 可全天时、全天候工作的独特优点，而且其三维成像能力避免了三维场景空间投影到二维平面造成的图像模糊问题。近场三维 SAR 成像技术可用于室内物品和人体的三维成像、隐藏物品探测、地质勘探、考古、城市建设、交通、军事等领域。项目针对近场三维SAR成像，主要开展了以下几个方面的研究：1.分析近场三维SAR成像机理并建立近场三维SAR成像模型；2.近场三维SAR成像处理算法；3.近场三维SAR成像仿真验证。为了获得三维雷达图像，宽带雷达运动产生的合成孔径须是二维的，即在高度向和方位向都形成合成孔径。结合信号接收方式和雷达收发方式考虑，圆周合成孔径雷达可以在不增加雷达接收复杂度的条件下实现目标区域的三维成像。项目在分析圆周合成孔径近场回波模型基础上，提出了基于球面波分解的圆周合成孔径三维成像算法。利用球面波分解理论，将近场条件下球面波转化为平面波的叠加，降低了算法复杂度。同时，从目标重建的推导过程中分析了该算法与基于信号相干累加的后向投影算法、聚焦因子算法的异同点。基于算法分析开展了仿真验证，结果表明，三种算法都可以对近场目标进行较好的三维重构，另外，聚焦因子算法可以获得最低的峰值旁瓣比和最高的成像分辨率；而基于球面波分解的成像算法所需运算时间最短。通过开展近场三维SAR成像研究，有助于推动近场成像技术的发展，对该技术进入实用阶段起到了促进作用。