APP下载

面向对象的高分辨率遥感影像建筑物震害信息提取

2016-01-05甘甜,李金平,李小强等

测绘工程 2015年4期

面向对象的高分辨率遥感影像建筑物震害信息提取

甘甜1,2,李金平1,2,李小强1,2,王刘伟1,2(1.云南师范大学旅游与地理科学学院,云南昆明650500;2.西部资源环境地理信息技术教育部工程研究中心,云南昆明650500)

摘要:建筑物的倒损信息是震后灾害评估的一项重要指标。文中应用震后高分辨率遥感影像数据,采用面向对象分类方法,以最优分割参数对影像进行分割,构建多尺度影像对象层次结构。通过影像对象的光谱、形状、纹理等特征及空间拓扑关系建立分类规则库,提取基本完好、受损和完全倒塌三类破坏等级的建筑物震害信息。结果表明,面向对象分类方法能够实现提取三类等级的建筑物震害信息,从而满足地震灾害快速评估要求。

关键词:面向对象;高分辨率遥感;震害信息提取;多尺度分割

中图分类号:P237文献标志码:A

收稿日期:2015-01-23;修回日期:2015-03-28

基金项目:云南省科技计划项目(2011XXL004)

作者简介:甘甜(1989-),女,硕士研究生.

Object-oriented method of building damage extraction from high-resolution images

GAN Tian1,2, LI Jin-ping1,2, LI Xiao-qiang1,2, WANG Liu-wei1,2(1.School of Tourism and Geographic Science, Yunnan Normal University, Kunming 650500, China; 2.GIS Technology Research Center of Resource and Environment in Western China of Ministry of Education, Kunming 650500, China)

Abstract:Building of damage information is an important index for earthquake disaster assessment.In this paper, the object-oriented classification method is used, based on the optimal parameters for segmentation of the image segmentation, by establishing image objects of multi-scale hierarchical structure, according to the spectrum feature, shape feature, texture feature and spatial topological relations to establish the rule database, negligible to slight damage, damage and collapsed completely of building damage information being extracted from high resolution remote sensing image data after the earthquake. The results show that object-oriented classification method can extract three-level building damage information, and to meet the requirements of rapid earthquake disaster assessment.

Key words:object-oriented; high-resolution remote sensing; earthquake damage extraction; multi-resolution segmentation

地震发生后快速做好震害调查与评价工作,是迅速实施灾后救援、重建以及最大限度降低经济财产损失的重要途径[1-2]。随着遥感技术的快速发展,不同平台和不同传感器的高空间分辨率遥感为地震灾害信息提取和震害快速评估提供了海量的、可靠的震前和震后遥感数据源,这些遥感数据可以节省人力、物力、财力并较大程度地提供大范围震害信息,在地震灾害损失快速评估中得到广泛应用[3-5]。

震后城镇建筑物震害信息的自动识别与分类是遥感震害调查中的关键技术,震害信息的提取精度直接影响地震灾害损失评估的结果[6]。……

登录APP查看全文