多源遥感数据建筑物震害信息联合分析
发布时间：2018-11-16 13:42
【摘要】：地震是造成人民生命和财产损失的最严重的自然灾害之一,地震灾害中建筑物的破坏信息对救灾和重建工作至关重要。遥感技术能够客观高效地获取大范围灾情信息,已经成为地震应急以及震害信息调查工作最重要的信息获取途径。目前,用于获取灾情信息的遥感传感器主要有两种,一是被动观测地物反射和辐射特性的光学传感器(光学遥感),另一种是主动发射微波的SAR传感器(雷达遥感)。高分辨率光学影像可用于详细评估地面灾情,但是不适用于云雪天气地区。而雷达遥感具有全天候全天时的特点,可以弥补光学遥感的不足。近年来,三维激光扫描技术(LiDAR)也被用于震后建筑物评估中,它可以获取地面精确的真三维点云,得到建筑物高程信息。不同的遥感数据源具有各自优势,将其联合起来可以更好地分析震害建筑物成像机理,提高震害建筑物提取精度。本文的主要研究内容是多源遥感数据建筑物震害信息联合分析与提取。建筑物震害信息在不同数据源中具有不同的表现特征,对其进行总结并定量分析可以有效地用于建筑物震害信息提取中；光学数据和LiDAR数据直观,可以用于辅助SAR图像解译,更好地对建筑物震害在SAR数据中的表现形式进行定量分析；多源数据特征丰富了面向对象提取的方法,并为震后快速评估提供了保障。主要开展的研究工作如下：(1)总结了震害建筑物在光学、SAR、LiDAR数据中的不同表现特征。以都江堰为试验区,建立房屋样本库,对特征参量进行统计,比较不同数据源统计结果的差异,为基于多特征融合的震害建筑物提取奠定基础。(2)展开了建筑物震害多源数据联合分析。根据SAR建筑物成像的不同几何模型,辅以光学数据、矢量和照片等,对样本库中建筑物SAR图像特征进行分析；对典型LiDAR点云建筑物进行剖面分析,对其形变进行定量研究,并分析其在对应SAR图像上的表现形式。(3)利用多源数据进行面向对象的建筑物震害信息提取：利用光学和LiDAR数据进行基于多特征融合的建筑物信息提取；利用震后光学、LiDAR数据分别对震后建筑物进行快速面向对象提取,获得优于单一提取方法精度的震后快速评估方法。
[Abstract]:Earthquake is one of the most serious natural disasters that cause the loss of people's lives and property. The damage information of buildings in earthquake disaster is very important for disaster relief and reconstruction. Remote sensing has become the most important information acquisition method for earthquake emergency and earthquake disaster information investigation, which can obtain disaster information objectively and efficiently. At present, there are mainly two kinds of remote sensing sensors for obtaining disaster information, one is optical sensor (optical remote sensing) for passive observation of reflectance and radiation characteristics of ground objects, the other is SAR sensor (radar remote sensing) for active microwave emission. High-resolution optical images can be used to assess ground damage in detail, but not in cloud and snow weather areas. Radar remote sensing has the characteristics of all-weather and all-day, which can make up for the deficiency of optical remote sensing. In recent years, 3D laser scanning technology (LiDAR) has also been used in post-earthquake building assessment, which can obtain accurate 3D point cloud and obtain building elevation information. Different remote sensing data sources have their own advantages. Combining them can better analyze the imaging mechanism of seismic damage buildings and improve the extraction accuracy of seismic damage buildings. The main research content of this paper is the joint analysis and extraction of earthquake damage information of multi-source remote sensing data buildings. The information of building earthquake damage has different characteristics in different data sources, which can be effectively used to extract the information of building earthquake damage by summarizing and quantitatively analyzing it. The optical data and the LiDAR data are intuitionistic and can be used to aid the interpretation of SAR images and to better quantitatively analyze the manifestation of building earthquake damage in SAR data. The feature of multi-source data enriches the method of object-oriented extraction and provides a guarantee for rapid assessment after earthquake. The main research works are as follows: (1) the different characteristics of seismic damage buildings in optical and SAR,LiDAR data are summarized. Taking Dujiangyan as the experimental area, the house sample bank is established, the characteristic parameters are counted, and the statistical results of different data sources are compared. It lays a foundation for the extraction of earthquake damage building based on multi-feature fusion. (2) the joint analysis of multi-source data of building damage is carried out. According to the different geometric models of SAR building imaging, with the help of optical data, vector and photo, the features of building SAR images in the sample database are analyzed. The typical LiDAR point cloud buildings are analyzed and their deformation is studied quantitatively. And analyze its representation on the corresponding SAR image. (3) using multi-source data to extract the information of building damage; using optical and LiDAR data to extract building information based on multi-feature fusion; Based on the post-earthquake optics and LiDAR data, a fast post-earthquake evaluation method is obtained, which is better than the single extraction method in fast object-oriented extraction of buildings after the earthquake.
【学位授予单位】：中国地震局地壳应力研究所
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P315.9
本文编号：2335661
[Abstract]:Earthquake is one of the most serious natural disasters that cause the loss of people's lives and property. The damage information of buildings in earthquake disaster is very important for disaster relief and reconstruction. Remote sensing has become the most important information acquisition method for earthquake emergency and earthquake disaster information investigation, which can obtain disaster information objectively and efficiently. At present, there are mainly two kinds of remote sensing sensors for obtaining disaster information, one is optical sensor (optical remote sensing) for passive observation of reflectance and radiation characteristics of ground objects, the other is SAR sensor (radar remote sensing) for active microwave emission. High-resolution optical images can be used to assess ground damage in detail, but not in cloud and snow weather areas. Radar remote sensing has the characteristics of all-weather and all-day, which can make up for the deficiency of optical remote sensing. In recent years, 3D laser scanning technology (LiDAR) has also been used in post-earthquake building assessment, which can obtain accurate 3D point cloud and obtain building elevation information. Different remote sensing data sources have their own advantages. Combining them can better analyze the imaging mechanism of seismic damage buildings and improve the extraction accuracy of seismic damage buildings. The main research content of this paper is the joint analysis and extraction of earthquake damage information of multi-source remote sensing data buildings. The information of building earthquake damage has different characteristics in different data sources, which can be effectively used to extract the information of building earthquake damage by summarizing and quantitatively analyzing it. The optical data and the LiDAR data are intuitionistic and can be used to aid the interpretation of SAR images and to better quantitatively analyze the manifestation of building earthquake damage in SAR data. The feature of multi-source data enriches the method of object-oriented extraction and provides a guarantee for rapid assessment after earthquake. The main research works are as follows: (1) the different characteristics of seismic damage buildings in optical and SAR,LiDAR data are summarized. Taking Dujiangyan as the experimental area, the house sample bank is established, the characteristic parameters are counted, and the statistical results of different data sources are compared. It lays a foundation for the extraction of earthquake damage building based on multi-feature fusion. (2) the joint analysis of multi-source data of building damage is carried out. According to the different geometric models of SAR building imaging, with the help of optical data, vector and photo, the features of building SAR images in the sample database are analyzed. The typical LiDAR point cloud buildings are analyzed and their deformation is studied quantitatively. And analyze its representation on the corresponding SAR image. (3) using multi-source data to extract the information of building damage; using optical and LiDAR data to extract building information based on multi-feature fusion; Based on the post-earthquake optics and LiDAR data, a fast post-earthquake evaluation method is obtained, which is better than the single extraction method in fast object-oriented extraction of buildings after the earthquake.
【学位授予单位】：中国地震局地壳应力研究所
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P315.9
【参考文献】
相关期刊论文 前5条
1 刘云华;屈春燕;单新建;宋小刚;张桂芳;张国宏;;SAR遥感图像在汶川地震灾害识别中的应用[J];地震学报;2010年02期
2 ;Post-earthquake assessment of building damage degree using LiDAR data and imagery[J];Science in China(Series E:Technological Sciences);2008年S2期
3 PESARESI Martino;;An improved automatic detection method for earthquake-collapsed buildings from ADS40 image[J];Chinese Science Bulletin;2009年18期
4 王瑞雪,叶燎原;应用遥感技术进行地震灾害快速调查的研究[J];昆明理工大学学报(理工版);2003年04期
5 张景发,谢礼立,陶夏新;建筑物震害遥感图像的变化检测与震害评估[J];自然灾害学报;2002年02期
相关硕士学位论文 前3条
1 马立广;地面三维激光扫描测量技术研究[D];武汉大学;2005年
2 董秀军;三维激光扫描技术及其工程应用研究[D];成都理工大学;2007年
3 金鼎坚;高分辨率SAR影像建筑物震害信息提取方法研究[D];中国地震局地震预测研究所;2012年
，本文编号：2335661
本文链接：https://www.wllwen.com/kejilunwen/diqiudizhi/2335661.html
