摘要
无人机遥感其灵活高效的技术优势可为震后应急救援与地震损失评估提供有力的技术支撑。与此同时建筑物信息作为地震灾情快速评估的重要评估对象,其重要性不言而喻。本研究选取大连长海县城岭村作为研究区域,结合无人机影像的光谱、纹理、几何特性,采用监督分类、基于规则的面向对象、LSD直线检测法对区域内建筑物信息开展提取测试,通过分析结果、归纳流程在建筑物解译精度和算法处理效能方面,讨论各项算法最佳适用环境和使用方式,为无人机遥感影像的建筑物自动提取提供合理的解决方案,为震后快速评估与辅助决策提供技术支持。
The flexible and efficient technical advantages of UAV remote sensing can provide strong technical support for post-earthquake emergency rescue and earthquake loss assessment.At the same time,building information is an important assessment object for rapid earthquake disaster assessment.Its importance is self-evident.This study selects Lingcun Village,Changhai County,Dalian as the research area,and combines the spectrum,texture and geometric characteristics of UAV images.The supervised classification,rule-based object-oriented and LSD linear detection method are adopted to extract and test the building information in the region.Through analysis results and induction process,the theoretical basis is provided for post-earthquake emergency rescue and auxiliary decision-making,and a reasonable solution is provided for automatic building extraction from unmanned aerial vehicle remote sensing images.
作者
于浩
徐汉卿
李雨佳
冯石
赵明朝
YU Hao;XV Han-qing;LI Yu-jia;FENG Shi;ZHAO Ming-zhao(Liaoning Earthquake Administration,Liaoning Shenyang110034,China;Liaoning Power Transmission and Substation Engineering CO.,LTD,Liaoning Shenyang 110024,China)
出处
《防灾减灾学报》
2020年第4期79-85,共7页
Journal of Disaster Prevention And Reduction
基金
辽宁省地震局青年基金“无人机遥感在地震应急中的探索与作用”(LZ201907)。
关键词
UAV
监督分类
面向对象
LSD直线检测
UAV
supervised classification
object oriented
LSD line detection
作者简介
于浩(1987-),男,辽宁省锦州市人,2012年毕业于吉林大学,本科,工程师,现主要从事地震应急指挥系统管理工作。
