SAR图像的地形形变信息提取算法研究与实现
ABSTRACT
Before the1950s,optical remote sensing is dominant in the field of remote sensing. For a long time,the imaging problem of optical remote sensing under harsh conditions has troubled many experts and scholars at home and abroad.Finding a remote sensing method adapting to any weather(alll-weather),24-hour observations(all-time)and having a strong penetrating power is the trend of The Times,synthetic aperture radar (SAR)was born with all these advantages.With the rapid development of radio astronomy,in the early90s,SAR combined radio astronomy interference technology, synthetic aperture radar interferometry(InSAR)technique appears,on this basis, differential synthetic aperture radar interferometry(D-InSAR)technique appears.Both InSAR technique and D-InSAR technique are using the phase information of SAR images to invert to obtain the surface information.The surface elevation information can be extracted by InSAR technique,and the surface deformation information can be extracted by D-InSAR technique.
This paper makes an important research on the key algorithm of terrain deformation information extraction for SAR images,and carries out all the software implementation.The premise of extracting deformation information from D-InSAR technology is the precise registration of SAR images.The accuracy of registration will directly affect the accuracy of elevation information and deformation information extraction.Therefore,the main research on registration has been done in this paper.The registration method studied in this paper is based on the features.Through the analysis of the edge features,corner features and feature matching criteria,the registration method based on the SIFT feature points is adopted so that the registration accuracy reaches the sub-pixel level and the entire algorithm is made into software.Based on this, other key algorithms for extracting terrain deformation information are studied.The interferogram is obtained by conjugate multiplication of the main and auxiliary images, and the median filter is used as the phase noise filtering method of this paper.Then, three kinds of mainstream methods for the flat-earth phase removal are studied.They are frequency shift method,flattening method based on rough precision DEM,and orbit parameter flattening method.In this paper,the orbit parameter flattening method is used to flatten and give out the main steps.Based on the flat-earth phase removal of
interferograms,two kinds of mainstream methods for phase unwrapping are studied, namely the minimal norm method and the path-following method.This paper adopts a phase unwrapping algorithm that uses the new quality map for path tracking.The main steps were analyzed and this method belongs to the second category.After phase unwrapping,the geometric features of InSAR and D-InSAR techniques are discussed, and the elevation and terrain deformation information are obtained through geometric expressions.In this paper,the method of extracting deformation based on the three-track method is adopted.From the registration of SAR images to the extraction of elevation and deformation information,all the programs in the relevant part are written using C++, and the interface is formed using Qt.Finally,software platform that can accurately extract elevation information and deformation information is formed.
Keywords:D-InSAR,Registration,Flat-earth Phase Removal,Phase Unwrapping, DEM,Terrain Deformation
目录
第一章绪论 (1)
1.1选题背景 (1)
1.2D-InSAR提取形变信息研究意义 (1)
1.3国内外研究现状和发展态势 (3)
1.4主要研究内容 (4)
1.5论文结构 (5)
第二章SAR图像精配准研究 (6)
2.1SAR基本原理和特点 (7)
2.1.1SAR基本原理 (7)
2.1.2SAR图像特点 (9)
2.2SAR图像的特征提取和特征匹配 (10)
2.2.1SAR图像的特征提取 (10)
2.2.2SAR图像的特征匹配 (17)
2.3基于SIFT特征点的精配准 (20)
2.3.1SIFT算法基本原理 (20)
2.3.2配准软件的实现与验证 (28)
2.4本章小结 (33)
第三章地形形变信息提取的关键算法 (34)
3.1InSAR和D-InSAR基本原理 (34)
3.1.1InSAR基本原理 (34)
3.1.2D-InSAR基本原理 (36)
3.2去除平地效应算法 (37)
3.3相位解缠算法 (40)
3.4高程及形变信息提取算法 (43)
3.4.1高程提取算法 (43)
3.4.2形变提取算法 (45)
3.5本章小结 (46)
第四章地形形变信息提取软件实现 (47)
4.1软件开发环境和软件成品 (47)
4.2软件主要功能 (48)
4.3软件主要模块 (50)
4.4软件数据源 (57)
4.5高程信息提取结果 (57)
4.6形变信息提取结果 (61)
4.7本章小结 (63)
第五章总结与展望 (64)
5.1总结 (64)
5.2展望 (64)
致谢 (66)
参考文献 (67)
在学期间取得的与学位论文相关的研究成果 (71)