基于面向对象分类的建筑物倒塌率计算方法研究  

Research on the Ratio of Collapsed Buildings Based on Object-Oriented Classification

作　　者：张磊[1] 张景发[2] 

机构地区：[1]天津市地震局,天津300201 [2]中国地震局地壳应力研究所,北京100085

出　　处：《地震》2009年第B10期139-145,共7页Earthquake

基　　金：“十一五”国家科技支撑计划（2006BAC13B03-0102）

摘　　要：建筑物倒塌是造成地震人员伤亡的主要原因，对地震应急救援与决策具有重要的指导意义。遥感以其综合、宏观、快速、动态的特点，为地震灾害信息调查和震害快速评估提供了一种可靠的信息源。面向对象的分类方法是针对高分辨率影像的一种新的分类方法。本文在总结遥感影像建筑物震害信息提取方法进展的基础上，将建筑物分为基本完好和毁坏两个等级。选取唐山地震和汶川地震的震后航片．利用面向对象的分类方法来识别影像上基本完好的建筑物，计算评估区内的建筑物倒塌率。实验结果表明，面向对象的分类方法充分利用了图像上的光谱、形状、纹理等特征，在很大程度上克服了基于像元分类方法的局限性，并且基于对象的建筑物倒塌率计算也取得了较好的精度。Collapse of buildings is the primarily cause for casualties in an earthquake, which is important in rescue and decision. Remote Sensing is a reliable information source for seismic investigation and evaluation using, with integrative, rapid, macroscopical and dynamic characteristics. Object-oriented classification is a new method to classify high-resolution images. Based on the development of building seismic information extraction methods, buildings are classified into two grades, slight to negligible damage building and collapsed building. Post-earthquake aerial photographs were selected for Tangshan Earthquake and Wenchuan Earthquake in this paper. Object-oriented classification was used to recognize slight to negligible damage buildings and compute the ratio of collapsed buildings. The result indicates that object-oriented classification makes full use of spectral, shape and texture information on images and images after classification are more integrated, Moreover, the precision of ratio of collapsed buildings is better using Object-koriented classification.

关 键 词：遥感 倒塌率 高分辨率遥感影像 面向对象的分类方法 

分 类 号：P315[天文地球—地震学]