基于无人机影像的长宁地震建筑物震害提取与成因分析

2019年6月17日,四川省长宁县发生6.0级地震,对当地建筑造成严重的破坏。选取震中长宁县双河镇为研究区,通过获取地震前后的无人机遥感数据,建立无人机遥感解译标志,对地震灾区的建筑物受损情况进行解译。在解译基础上进行无人机遥感建筑物震害提取,获取研究区的灾情信息。将目视解译方法得到的结果和无人机低空视频拍摄的震害结果进行混淆矩阵对比分析,得到两者总体精度(81%),kappa系数为0.74,具有较高的一致性。分析结果显示:研究区域内共解译建筑1867栋,损坏、部分倒塌和倒塌的建筑物占研究区总建筑物的64.06%,其中倒塌房屋为年代久远土木结构房屋,集中在研究区的东北部分(9.27%),另外,基本完好的建筑物(占比36.94%)主要位于研究区西部,为近年来新修的建筑,抗震性能良好。

UAV and GIS Based Structure Damage Assessment after Changning Earthquake

On June 17,2019,a Ms 6.0 earthquake struck Changning County,Sichuan Province,China.It induced severe damage to local buildings.In this study,we selected Shuanghe town,the epicenter of Changning earthquake,as the study area,where the damage of buildings and infrastructures were the most.Unmanned Aerial Vehicle(UAV)observation data before and after the earthquake were utilized to visually interpret the information of damage buildings,with the help of two kinds of data,DOM and video.The overall accuracy is 81%and the kappa coefficient is 0.74,which has high consistency.The damage of buildings induced by the earthquake was extracted to assess the disaster losses in Shuanghe town.According to the results,1,867 buildings were interpreted.Damaged,partially collapsed and collapsed buildings accounted for 64.06%,of which the collapsed houses took up 9.27%,which were old civil structures that were mainly located in the northeast part of the study area.While intact buildings,which accounted for 35.94%,were mainly located in the west of Shuanghe.New buildings built in recent years had good seismic performance.