D-INSAR技术在滑坡信息提取中的应用研究

发布时间：2018-04-20 12:45

本文选题：干涉图 + 去平地　；参考：《电子科技大学》2013年硕士论文

【摘要】：滑坡是一种破坏力很大的常见自然地质灾害，其能够引起泥石流以及山体垮塌，会造成人民生命财产安全及国家建设的巨大损失。因此对滑坡进行监测，是减少滑坡灾害危害性的有效方式。随着干涉合成孔径雷达技术（InSAR）的发展，，合成孔径雷达差分干涉测量（D-InSAR）作为其扩展应用，具有宽覆盖率、高空间分辨率、全天时、全天候、大区域性、高灵敏度等特点，且对地表的微小形变具有厘米级别甚至更小尺寸精度的探测能力，可以获取高精度、高分辨率的地面高程信息，并通过滑坡灾害前后两幅干涉图像的比对，达到精密测定滑坡前后的地表形变的效果，因此本文采用D-InSAR技术应用于小型滑坡的形变信息提取研究。本论文系统地阐述了合成孔径雷达差分干涉技术应用于滑坡信息提取的发展历程、应用现状、SAR成像原理以及INSAR基本原理，并在此基础上深入应用研究了D-INSAR技术主要处理流程，并对其中的主要算法做了深入讨论与分析，并运用编程软件进行编程以实现主要算法的功能。主要研究内容包括以下几个部分：将配准后的两幅主、辅图像形成干涉条纹图、将干涉条纹图进行平地效应去除处理、相位噪声滤波、相位解缠、高程转换及形变提取。并根据RADARSAT-2提供的SAR数据对D-INSAR技术的主要算法进行了数据处理及算法验证。其中基于轨道参数的去平地效应方法，直接拟合卫星轨道方程，避免多次迭代的复杂运算，运算效率得以提高。并对几种滤波方法进行残差点数目、信噪比、条纹边缘保持完整性进行分析，得出自适应中值滤波法具有较好滤波效果的结论。且采用路径跟踪法中的枝切法进行了相位解缠处理，该算法优点在于运行时间短。最后对具有滑坡信息的SAR图像进行了D-INSAR处理提取了滑坡的地表形变信息，结果表明D-INSAR技术能提取小型滑坡厘米级别的微小形变信息。
[Abstract]:Landslide is a kind of common natural geological disaster with great destructive power. It can cause debris flow and collapse of mountain body, which will cause great loss of people's life and property safety and national construction. Therefore, monitoring landslide is an effective way to reduce the hazard of landslide. With the development of Interferometric synthetic Aperture Radar (InSAR) technology, D-InSAR (differential Interferometry of synthetic Aperture Radar) as its extended application has the characteristics of wide coverage, high spatial resolution, all-weather, all-weather, large area, high sensitivity and so on. And the small deformation of the surface has the ability to detect the precision of centimeter level or even smaller size, which can obtain the high precision and high resolution information of the ground elevation, and compare the two interference images before and after the landslide disaster. Therefore, D-InSAR technique is used to extract deformation information of small landslide. In this paper, the development of differential synthetic aperture radar (SAR) interferometry in landslide information extraction, the principle of SAR imaging and the basic principle of INSAR are systematically described. Based on this, the main processing flow of D-INSAR technology is deeply studied. The main algorithms are discussed and analyzed in depth, and programming software is used to realize the functions of the main algorithms. The main contents of this paper are as follows: the two principal and auxiliary images after registration are used to form interference fringes, the interference fringes are processed by leveling effect removal, phase noise filtering, phase unwrapping, height conversion and deformation extraction. According to the SAR data provided by RADARSAT-2, the main algorithms of D-INSAR technology are processed and verified. Among them, the deflattened earth effect method based on orbit parameters can directly fit the satellite orbit equation, avoid the complex operation of multiple iterations, and improve the operation efficiency. The number of residual handicap, SNR and fringe edge integrity of several filtering methods are analyzed, and the conclusion that adaptive median filtering method has better filtering effect is obtained. The method of branch cutting in path tracking is used to unwrap the phase, and the advantage of the algorithm is that the running time is short. Finally, the SAR image with landslide information is processed with D-INSAR to extract the ground deformation information of the landslide. The results show that the D-INSAR technique can extract the small deformation information of the small landslide in centimeter level.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P642.22;P225

【参考文献】