【作者】 汪鲁才；

【导师】 王耀南；

【作者基本信息】 湖南大学 ， 模式识别与智能系统， 2006， 博士

【摘要】 干涉合成孔径雷达(InSAR)成像技术是在合成孔径雷达(SAR)技术基础上发展起来的雷达成像技术。它既继承了SAR的全天候、大范围、有一定穿透能力等优点，又具有对目标三维地貌进行高精度成像、对目标地面的慢速动目标和地壳微小移动作高精度检测等的能力。因此，该技术在军事、国民经济建设中，有着极其广泛的应用。 论文首先介绍了星载合成孔径雷达干涉测量三维成像技术的发展历程，简述了国内外有关干涉合成孔径雷达发展现状。论述了干涉合成孔径雷达三维成像的基本原理。分析了合成孔径雷达干涉测量系统模型，说明了星载InSAR获取地形信息的数据处理流程和合成孔径雷达差分干涉测量技术。 论文在研究了干涉数据图像进行相关配准、最大谱配准方法的基础上，提出了一种将相关配准与最大谱配准相结合的配准方法。在满足配准要求的条件下，减少了配准计算的工作量，并通过截取真实SAR数据进行实验，结果验证了该算法的有效性。论文分析了配准误差来源，给出了相干系数和干涉相位方差、干涉相位的概率密度分布之间的关系，以有效衡量两幅SAR图像的配准精确度。 论文论述了InSAR成像技术中干涉纹图的平地效应形成的原因，在简述消除平地效应常用方法的基础上，基于时域内乘以复指数的原理，提出了一种采用多次样条插值算法求取精确的平均空间频率，较好地达到消除平地效应的目的。实验表明：该算法与常用的消除平地效应方法相比较，能够更好地减少干涉相位图中的残差点。 论文分析了干涉纹图的噪声来源和干涉纹图的特性。根据常用的圆周中值滤波方法和圆周均值滤波方法，从小波变换理论出发，将小波分析和小波包分析应用于干涉纹图的滤波中，提出了一种基于小波分析和中值滤波相结合的滤波方法。并利用该方法对实际的干涉纹图数据进行实验，给出了将小波分析和中值滤波相结合方法用于干涉纹图滤波和小波包分析用于干涉纹图滤波的实验结果。结果表明该方法在减少残差点和保持干涉相位图的边缘细节方面具有有效性。 论文研究了经典的分支切割相位解缠算法和最小二乘相位解缠算法。在分析经典的分枝切割相位解缠方法的基础上，针对分枝切割算法的特点，提出了一种旨在加快形成分枝切割线速度的基于映射的分枝切割相位解缠的改进算法。实验证明了该算法形成分枝切割线所需计算时间比采用经典的分枝切割相位解缠方法形成分枝切割线所需时间大大减少，而两种方法的解缠效果基本相同。同时，论文还分析了获取相对数字高程模型(DEM)的方法。更多还原

【Abstract】 Interferometric Synthetic Aperture Radar(InSAR) image technique is a new radar image technique developed from Synthetic Aperture Radar (SAR).Carried forward the merit of SAR, InSAR can create high resolution DEM and accomplish change detection, such as slow-moving targets or subtle shift of the earth’s surface. It can be applied to many areas of military and civil.The history developing of InSAR is reviewed, and the actual state of InSAR is depicted briefly in the thesis. The basic principle of InSAR imaging technique is introduced in the thesis firstly. Then, the system model of InSAR is analyzed. The flow of data processing for InSAR is discussed, and the differential InSAR is introduced in the thesis.The InSAR complex images registration methods of correlation matching and spectral maximum matching are discussed in the thesis, and a new method of InSAR complex images registration using correlation matching combined with spectral maximum matching is presented. Compared with other method, the merit of this new method is less time spending on calculating for complex images registration on the condition that is up to complex images registration precision, and the results of experiment show that this new method was resultful. The error sources of complex images registration is analyzed, and the relation between the correlation coefficient, the differential of interferometric phase and the PDF.The reasons to form the horizontal ground effect of interference pattern in InSAR imaging technique were discussed in this thesis, and the common methods of eliminating the horizontal ground effect were explained suitably. One method was to multiply the phase signal of interference pattern by a complex exponent changed with the time, and the other one was moving the center of frequency spectral to the zero frequency spectral. A new algorithm to obtain a precision frequency based on linearity interpolation technique was advanced, and the results of experiment show that the algorithm was effective.By analyzing the noise sources and the characteristic of interferogram phase, the average-value-filter and the median-value-filter are employed usually in interferogram phase filter. An algorithm of interferometric phase filter of InSAR based on wavelet analysis filter algorithm combined with median filter algorithm is advanced to更多还原