摘要
为了快速地确定地震等自然灾害引起的受灾区域范围,并对其受灾程度进行及时评估,本文采用面向对象的建筑物检测方法,基于高分辨率遥感影像所包含的地物几何结构和纹理特征信息,提出了一种建筑物震害信息提取与评估的方法和技术流程.在此基础上,以2010年玉树MS7.1地震部分地区地震前后的QuickBird影像为例,对受灾区域震前、震后建筑物的形状、面积等信息进行提取,提取精度分别为88.53%和90.21%,对该区域建筑物变化信息进行提取所获取的建筑物变化信息精度为79.68%,统计变化区域像素个数,确定变化面积为15 923.52m2,占研究区域总面积的68.16%,因此评估其为中重度受灾区域.本文结果与实地考察结果一致,证实了这种快速的震害信息提取与评估流程切实有效,能够快速评估受灾区,为灾后第一时间抢险及救援提供重要参考.
In order to rapidly determine the scope of stricken area and timely assess the extent of damages after an earthquake,this paper proposes a technical process of rapid extracting and evaluating building damage information by using the geometric structure and texture feature information of high resolutionremote sensing images based on the object-oriented building detection method.The process can rapidly locate the disaster areas,which is of great significance to the post-disaster first opportunity rescue.Taking the Yushu area as an example,buildings of disaster area are extracted based on the QuickBird images before and after the Yushu earthquake,and the extraction precisions of buildings is88.53%and 90.21%,respectively.The extraction accuracy of building changing information is 79.68%,and changing area reaches 15 923.52m2,which accounts for 68.16% of the entire studied area,therefore the area is evaluated as moderatelysevere disaster area.The results of this paper are consistent with those of the field investigations,proving that the rapid seismic damage information extraction and evaluation process is effective.The presented method can quickly estimate the disaster areas,and provide an important reference for the first time rescue.
作者
赵妍
张景发
姚磊华
Zhao Yan Zhang Jingfa Yao Leihua(） College of Engineering and Technology, China University of Geosciences （ Beijing） , Beijing 100083, China ） Institute of Crustal Dynamics, China Earthquake Administration, Beijing 100085, China)
出处
《地震学报》
CSCD
北大核心
2016年第6期942-951,共10页
Acta Seismologica Sinica
基金
国家自然科学基金(41374050)
高分遥感地震监测与应急应用示范系统(一期)项目(31-Y30B09-9001-13/15)共同资助
关键词
高分辨率遥感
震害信息提取与评估
面向对象变化检测
high resolution remote sensing
building damage extraction and evaluation
object-oriented change detection
