极化干涉合成孔径雷达与层析成像技术研究

发布时间：2019-06-18 15:09

【摘要】：极化干涉合成孔径雷达(SAR)综合了极化信息与传统干涉信息感知能力,因其具备了全天时、全天候、远距离对目标电磁特征、空间位置、空间结构探测和环境监测的能力,在地形测绘、海洋监测、农林业测绘、军事侦察、气象制图、环境保护以及灾情监测等各个应用领域中表现出了巨大优势与潜力,已成为现代军事和民用遥感信息获取的重要手段。在单基线极化干涉SAR应用中,散射过程的模型失配极大地限制了地表地物信息的获取能力。极化干涉SAR技术的实用化要求对实际观测数据与雷达系统参数、场景特征参数之间的关联进行深入的分析与理解,还需要通过信号处理方法进一步提升地物参数反演精度。借助于海量的对地遥感观测数据,沿垂直航向空间观测通道的增加大大丰富了极化干涉SAR所能获取的目标信息量。多基线极化干涉SAR技术和层析SAR成像技术能够有效获取自然场景垂直结构的散射特征、实现复杂场景的三维重构,是现代成像雷达发展的一个重要方向。然而,地物目标对电磁信号的散射过程极为复杂,多基线极化干涉观测数据与目标散射特征之间的映射法则难以确定。在三维重构方面,层析三维成像受制于最优基线配置及实现的复杂性以及传统二维SAR成像处理误差的影响,其层析维聚焦性能的提升亟需高效稳健的高精度聚焦方法。针对这些问题,本论文旨在通过信号处理方法提高极化干涉SAR、层析SAR对地遥感探测信息获取的精度与稳健性,研究内容主要针对地表散射模型失配问题、单基线极化干涉SAR地表参数反演、双基线极化干涉分布模型构建与体散射解模糊、层析SAR基线优化设计与三维成像四个关键点。本文围绕×××预先研究项目“×××信息获取技术”、“×××载荷技术”以及×××探测技术项目“×××技术与实验研究”等项目研究任务,对极化干涉SAR地表参数反演、极化层析SAR成像进行了研究。全文内容主要围绕地表散射模型失配问题、单基线极化干涉SAR地表参数反演、双基线体散射解模糊和层析SAR基线优化设计与三维成像进行展开,主要内容可概括为以下四个部分:第一部分研究基于雷达系统参数和场景特征参数的地表散射模型构建。引出了散射过程的模型失配问题,分析了观测数据与雷达系统参数、场景特征参数(包括植被参数和地形坡度参数)之间的关联,确定了地形坡度参数对基于地表散射模型的植被参数反演结果的影响方式,建立了极化干涉信号与雷达系统及场景特征参数之间的映射关系,在模型级上解决了由地形起伏引起的模型失配问题。第二部分研究单基线极化干涉SAR地表参数反演信号处理方法。给出了植被覆盖下地形干涉相位的估计方法,提出了基于数据驱动的极化干涉SAR数据幅度、相位非平稳特性自适应校正方法。针对观测数据的非均匀分布特性,提出了映射空间均衡化校正技术。分析了以单一极化通道数据先验信息作为参考的失效情况,首次提出了基于地体散射比差异最大化的先验信息融合技术,并引入主成分分析方法,提高了线性变化模型特征提取的精度和稳健性。第三部分研究双基线极化干涉SAR地表参数反演方法。提出了双基线极化干涉SAR分布模型的构建方法,分析了双基线极化干涉SAR解决体散射估计模糊的基本原理,通过仿真试验分析方法验证了解模糊方法的有效性。分析了双基线观测数据的聚类现象,提出了基于最大一致性准则的双基线地表参数反演算法,并利用仿真数据对不同的长短基线比例条件下的算法性能和误差容忍度进行了分析与评估,该方法可以有效地解决体散射估计的模糊问题。第四部分研究极化层析SAR基线优化设计与三维成像。针对非均匀线性阵列阵元位置的优化设计,分析并推导了机载层析SAR系统阵列方向图的近场扫向不变性和阵元位置优化模型,提出了一种目标函数离散栅格化方法,将原问题近似为易于求解的极小极大最优化问题。针对自然场景的三维重构,提出了基于最优基线配置的全极化层析SAR三维成像处理算法框架,利用正射影像校正技术得到了大地坐标下的三维成像结果,实现了基于酉变换的全极化层析三维信息表达。
[Abstract]:The polarization interference synthetic aperture radar (SAR) combines the polarization information and the traditional interference information sensing capability, and has the capability of all-weather, all-weather, long-distance to the target electromagnetic characteristic, the space position, the spatial structure detection and the environmental monitoring, The great advantages and potential of agricultural and forestry mapping, military reconnaissance, meteorological mapping, environmental protection and disaster monitoring have become an important means of information acquisition in modern military and civil engineering. In the application of single-base-line polarization interference SAR, the model mismatch of the scattering process greatly limits the acquisition ability of surface figure information. The practical requirement of the polarization interference SAR technology is to carry out in-depth analysis and understanding of the correlation between the actual observation data and the parameters of the radar system and the characteristic parameters of the scene, and further improve the inversion accuracy of the figure parameters by the signal processing method. By means of mass-to-ground remote sensing observation data, the increase of the observation channel along the vertical heading space greatly enriches the target amount of information that the polarization interference SAR can obtain. The multi-baseline polarization interference SAR technology and the chromatographic SAR imaging technology can effectively acquire the scattering characteristics of the vertical structure of the natural scene, realize the three-dimensional reconstruction of the complex scene, and is an important direction of the development of the modern imaging radar. However, the object-to-ground object is very complicated in the scattering process of the electromagnetic signal, and the mapping rule between the multi-baseline polarization interference observation data and the target scattering feature is difficult to determine. In the aspect of three-dimensional reconstruction, the three-dimensional imaging is subject to the optimal baseline configuration and the complexity of implementation as well as the influence of the traditional two-dimensional SAR imaging processing error, and the enhancement of the chromatographic-dimensional focusing performance of the three-dimensional imaging is in urgent need of a highly efficient and robust high-precision focusing method. In order to solve these problems, the purpose of this paper is to improve the accuracy and robustness of the polarimetric interference SAR and the ground-ground remote sensing information acquisition by the signal processing method. The research content is mainly focused on the surface scattering model mismatch problem, single-base-line polarization interference SAR surface parameter inversion, The dual-baseline polarization interference distribution model is used to construct the four key points of the body-scattering and deblurring, the base-line optimization design of the chromatographic SAR and the three-dimensional imaging. In this paper, the research on the surface parameter inversion and polarization-tomography (SAR) imaging of the polarization interference SAR is studied in this paper. The paper mainly focuses on the surface scattering model mismatch problem, single-base-line polarization interference (SAR) surface parameter inversion, double-base-baseline-body-scattering-deblurring, and tomographic-SAR baseline optimization design and three-dimensional imaging. The main contents can be summarized as follows: The first part studies the construction of the surface scattering model based on the radar system parameters and the scene characteristic parameters. The model mismatch problem of the scattering process is introduced, the correlation between the observation data and the radar system parameters, the scene characteristic parameters (including the vegetation parameters and the terrain slope parameters) is analyzed, the influence of the terrain slope parameters on the vegetation parameter inversion result based on the surface scattering model is determined, The mapping relation between the polarization interference signal and the radar system and the scene characteristic parameters is established, and the model mismatch problem caused by the relief of the terrain is solved at the model level. The second part is to study the surface parameters inversion signal processing method of single-line polarization interference SAR (SAR). In this paper, the estimation method of the terrain interference phase under the vegetation cover is given, and the self-adaptive correction method for the amplitude and phase non-stationary characteristics of the polarization interference SAR data driven by the data is proposed. In view of the non-uniform distribution of the observed data, a mapping space equalization correction technique is proposed. In this paper, a prior information fusion technique based on the maximization of the dispersion ratio of the ground body is proposed for the first time, and the accuracy and the robustness of the feature extraction of the linear change model are improved. In the third part, the method of surface parameter inversion of dual-baseline polarization interference SAR is studied. In this paper, a method for constructing a dual-baseline polarization interference SAR distribution model is proposed, and the basic principle of the double-baseline polarization interference SAR solution is analyzed, and the effectiveness of the fuzzy method is verified by means of the simulation test. The clustering phenomenon of double-baseline observation data is analyzed, a double-baseline surface parameter inversion algorithm based on the maximum consistency criterion is proposed, and the algorithm performance and error tolerance are analyzed and evaluated by using the simulation data. The method can effectively solve the fuzzy problem of the body scattering estimation. The fourth part studies the design of the baseline optimization and the three-dimensional imaging of the polarization-chromatographic SAR. According to the optimal design of the array element position of the non-uniform linear array, the near-field scanning direction invariance and the array element position optimization model of the array direction map of the airborne chromatographic SAR system are analyzed and derived, and a method for discretizing a target function is presented. The problem of the original problem is approximated as a very small maximum optimization problem which is easy to solve. Aiming at the three-dimensional reconstruction of the natural scene, a three-dimensional imaging processing algorithm framework based on the optimal baseline configuration is proposed, and the three-dimensional imaging result under the geodetic coordinate is obtained by using the orthophoto-image correction technology, and the three-dimensional information expression of the full-polarization chromatography based on the optimal baseline configuration is realized.
【学位授予单位】：西安电子科技大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN957.52

【相似文献】