迭代高精度InSAR相位解缠方法研究

发布时间：2018-03-27 10:30

本文选题：InSAR　切入点：相位解缠　出处：《电子科技大学》2017年硕士论文

【摘要】：干涉合成孔径雷达(InSAR)除有全天时、全天候等独特优点外,还可获取目标的高程信息,越来越广泛地被应用到大范围地形高程测绘中,其中,相位解缠技术起着至关重要的作用。基于全局类解缠算法整体解缠精度高的优势,本文针对其中的四次快速傅里叶变换(4-FFT)算法进行了深入研究并提出改进,并从原理上将之扩展到多基线,提出基于多基线的高精度4-FFT算法,最后结合仿真及实测数据对比分析了该算法性能。具体工作及创新如下:(1)阐述了合成孔径雷达干涉测量技术的基本原理。首先介绍了InSAR测高基本原理,给出了InSAR数据处理流程,然后阐述了相位解缠基本原理,接着对传统的相位解缠算法及其优缺点做了简要介绍,提出高精度解缠算法研究的必要性,最后给出了相位解缠算法评价体系,为后面评判解缠算法性能提供依据。(2)研究了四次FFT相位解缠算法。首先分析了4-FFT算法的原理,通过对真实相位与缠绕相位的关系式取拉普拉斯变换,借助真实相位的拉普拉斯变换与缠绕相位间的关系,进行四次FFT求得解缠相位,获得全局类高精度4-FFT算法。然后将4-FFT算法与同样具有全局类特点的最小二乘算法进行了原理上的分析比较,说明了4-FFT算法解缠精度更高,仿真与实测数据同样表明了该特性。(3)提出了一种迭代式的高精度4-FFT相位解缠算法。针对传统4-FFT相位解缠算法在低信噪比情况下解缠误差大甚至失效的问题,提出了一种基于迭代的高精度4-FFT相位解缠算法。该算法利用传统4-FFT算法获得初始解缠相位,与缠绕相位建立相位误差主值求解公式,对误差主值进行4-FFT解缠求得误差真实值,迭代补偿到初始解缠相位,不断减小解缠误差提高解缠精度。最后在不同信噪比下与传统4-FFT和最小二乘法进行了比较,仿真和实测数据表明,所提算法不仅在高信噪比下解缠性能优良,在低信噪比环境仍能保持高的解缠精度。(4)提出了迭代多基线高精度4-FFT相位解缠算法。针对单基线相位解缠不能解决地形高度突变引起的相位层叠和解缠精度不高等问题,提出了基于多基线的迭代高精度4-FFT相位解缠算法。结合基线长度与干涉相位的关系,利用其它基线相位丰富所求基线相位信息,将4-FFT算法扩展到多基线相位解缠,并通过误差迭代补偿技术进一步提高解缠精度。仿真及实测数据表明,其利用其它视角的相位信息融合本基线相位有效解决了相位层叠问题,并通过与高精度多基线最小二乘算法的比较,证明所提算法解缠精度更高。
[Abstract]:Interferometric synthetic Aperture Radar (InSAR) has the unique advantages of all-day, all-weather and so on, and it can also obtain the height information of the target, which is more and more widely used in the large-scale topographic elevation mapping. Phase unwrapping technique plays an important role. Based on the advantage of global unwrapping algorithm with high global unwrapping accuracy, this paper makes a thorough study on the 4-FFTs algorithm and proposes an improved algorithm. In principle, it is extended to multi-baselines, and a high-precision 4-FFT algorithm based on multi-baselines is proposed. Finally, the performance of the algorithm is analyzed by comparing the simulated and measured data. The main work and innovation are as follows: 1) the basic principle of synthetic Aperture Radar (SAR) interferometry is expounded. Firstly, the basic principle of InSAR altimetry is introduced, and the InSAR data processing flow is given. Then the basic principle of phase unwrapping is expounded, and then the traditional phase unwrapping algorithm and its advantages and disadvantages are briefly introduced, and the necessity of research on high precision unwrapping algorithm is put forward. Finally, the evaluation system of phase unwrapping algorithm is given. The fourth FFT phase unwrapping algorithm is studied in this paper. Firstly, the principle of 4-FFT algorithm is analyzed, and the Laplace transform is used to express the relationship between real phase and winding phase. With the help of the relation between the Laplace transformation of real phase and the phase of winding, the unwrapping phase is obtained by FFT. The global class high precision 4-FFT algorithm is obtained. Then the 4-FFT algorithm is compared with the least square algorithm which also has the global class characteristics. It shows that the 4-FFT algorithm has higher unwrapping precision. Simulation and actual data show that this property is the same.) an iterative 4-FFT phase unwrapping algorithm with high precision is proposed. The traditional 4-FFT phase unwrapping algorithm has a large unwrapping error or even a failure in the case of low signal-to-noise ratio (SNR). A high precision 4-FFT phase unwrapping algorithm based on iteration is proposed, which uses the traditional 4-FFT algorithm to obtain the initial unwrapping phase, and establishes a formula for calculating the principal value of the phase error with the winding phase. The real value of the error is obtained by the 4-FFT unwrapping of the error principal value. The iterative compensation to the initial unwrapping phase reduces the unwrapping error continuously and improves the unwrapping accuracy. Finally, compared with the traditional 4-FFT and the least square method at different SNR, the simulated and measured data show that, The proposed algorithm not only has good unwrapping performance under high signal-to-noise ratio (SNR), In this paper, an iterative multi-baseline high-precision 4-FFT phase unwrapping algorithm is proposed. The single-baseline phase unwrapping algorithm can not solve the problem of phase stacking and unwrapping caused by topographic height abrupt change. An iterative high-precision 4-FFT phase unwrapping algorithm based on multiple baselines is proposed. Considering the relationship between baseline length and interference phase, the 4-FFT algorithm is extended to multi-baseline phase unwrapping by using other baseline phases to enrich the desired baseline phase information. The error iterative compensation technique is used to further improve the unwrapping accuracy. The simulation and measured data show that the phase stacking problem is effectively solved by using the phase information fusion of other angles of view to fuse the baseline phase. Compared with the high-precision multi-baseline least squares algorithm, the proposed algorithm has higher unwrapping accuracy.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN957.52

【参考文献】