基于GAMMA的差分干涉测量算法研究

发布时间：2018-04-22 03:23

本文选题：GAMMA + 差分干涉测量　；参考：《山东科技大学》2017年硕士论文

【摘要】：差分干涉测量技术在我国起步较晚,但近三十多年来发展迅猛。从最初的应用于地震、火山活动的监测,到逐渐成熟和研究深入将研究重点转移至地面沉降、山体滑坡等细微的地表形变。地表形变与人们的生命财产生活安全息息相关,利用差分干涉测量技术监测区域地表沉降是一种新兴便捷有效的监测方法,优势突出,具有重要的研究意义。相较于传统的大地测量技术,差分干涉测量技术工作覆盖面积大、能全天时、全天候工作,受雨、雾等恶劣天气的影响小;穿透性强,能够穿透云层、烟尘、雪层甚至地表植被等传统光学传感器成像困难的地区获取地表信息;其成像不仅能够记录后向散射信息,还可记录相位信息,在地表信息提取方面表现出较大优势。主要研究内容包括:1)分析合成孔径雷达的成像原理,介绍合成孔径雷达图像的主要特征,阐述干涉测量技术中干涉相位的形成原理和干涉测量原理。2)针对D-InSAR中涉及的各阶段的算法,研究了单视复影像的配准、生成干涉图、基线估计、相位解缠和相位差分等的算法。利用GAMMA软件通过二轨法开展D-InSAR的应用及分析。3)针对二轨法、三轨法和四轨法这三种差分干涉测量技术的处理方法,阐述其各自的工作流程。利用拉斯维加斯数据,进行了不同方法结果的对比分析。针对三轨法处理不同像对的问题,着重对比分析了地形像对和差分像对在不同算法中的效率。4)基于一维、二维相位解缠原理,深入研究了三大类相位解缠算法:路径跟踪的相位解缠算法、基于最小范数法的相位解缠算法和基于网络规划的相位解缠算法。针对枝切树区域增长法和基于三角网的最小费用流相位解缠算法的,进行实验分析,并提出了一种基于MATLAB的分析相位解缠效果的评价方法,利用BAM数据验证了该评价方法的有效性,并应用于LasVegas地区。
[Abstract]:Differential interferometry technology started late in China, but it has developed rapidly in recent 30 years. From the initial application to the monitoring of earthquake and volcanic activity to the gradual maturation and deep research, the research focus will be shifted to ground subsidence, landslide and other minor surface deformation. Surface deformation is closely related to people's life and property life and security. Using differential interferometry to monitor regional surface subsidence is a new convenient and effective monitoring method with outstanding advantages and important research significance. Compared with traditional geodetic techniques, differential interferometry has a large area of coverage, can work all day, all weather, and is less affected by bad weather, such as rain and fog; it is highly penetrating and can penetrate clouds, smoke, dust, etc. The traditional optical sensors such as snow layer and surface vegetation are difficult to obtain surface information, which can not only record backscattering information, but also record phase information, showing a great advantage in surface information extraction. The main research contents include: (1) analyzing the imaging principle of synthetic Aperture Radar (SAR), introducing the main characteristics of SAR image, This paper describes the principle of interferometric phase formation and interferometric measurement in interferometry. (2) aiming at the algorithms of each phase involved in D-InSAR, the algorithms of registration, interferogram generation, baseline estimation, phase unwrapping and phase difference grading of single view complex image are studied. Using GAMMA software to develop the application and analysis of D-InSAR through the two-rail method. 3) aiming at the processing methods of the three differential interferometric techniques, the two-rail method, the three-track method and the four-track method, their respective workflows are expounded. By using Las Vegas data, the results of different methods are compared and analyzed. Aiming at the problem of processing different image pairs by the three-track method, the efficiency of topographic image pair and differential image pair in different algorithms is compared and analyzed. 4) based on the principle of one-dimensional and two-dimensional phase unwrapping, Three kinds of phase unwrapping algorithms are studied in detail: phase unwrapping algorithm based on path tracking, phase unwrapping algorithm based on minimum norm method and phase unwrapping algorithm based on network planning. In this paper, an experimental analysis is carried out on the dendritic tree region growth method and the minimum cost flow phase unwrapping algorithm based on triangulation, and an evaluation method based on MATLAB is proposed to analyze the phase unwrapping effect. The validity of the evaluation method is verified by BAM data and applied to LasVegas area.
【学位授予单位】：山东科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P642.26

【参考文献】