MATLAB软件与ALOS遥感影像在震害建筑物自动识别中的应用——以都江堰市区为例  

Application of MALAB Software and ALOS Remote Sensing Imagine to Automatic Recognition of Earthquake Damage Building:Take Dujiangyan Urban As an Example

作　　者：曾繁如[1,2] 何政伟[1,2] 李喆[1,3] 

机构地区：[1]成都理工大学地质灾害防治与地质环境保护国家重点实验室,四川成都610059 [2]成都理工大学数字国土与生态科学研究所,四川成都610059 [3]成都市国土资源局国土规划地籍事务中心,四川成都610000

出　　处：《西北地震学报》2009年第4期327-332,共6页Northwestern Seismological Journal

基　　金：国家863重点项目(2007AA120306)

摘　　要：利用遥感技术进行震害建筑物的自动识别可为震害的快速评估与救灾决策提供科学可靠的依据。本文从震害建筑物在高分辨率遥感影像下灰度的特征入手,以5.12汶川特大地震后都江堰市区ALOS遥感影像为数据源,在MATLAB平台下对影像进行灰度增强处理、数学形态学重构以及连接、填充处理,并结合区域统计特性最后自动识别震害房屋。结果表明,利用ALOS影像丰富的纹理特征及空间结构信息与MATLAB在数学形态学处理中的优势能够准确有效地提取震害建筑物信息。Remote sensing technology can provide a reliable scientific basis to the aspect of automatic recognition of earthquake damage building in the rapid evaluation and rescue decision. In this paper, based on the data of ALOS remote sensing imagine on Dujiangyan urban area after 5 ·12 Wenchuan great earthquake, taking the imagine characters of earthquake damage buildings from high resolution remote sensing as the starting point, using MATLAB software to processings of gray enhancemen, mathematic morphology reconstruction, bridge and filling, the damage buildings are automatically recognited. The result shows that it can accurately extract the earthquake damage building information by using abundant textural property and spatial structure information of ALOS remote sensing imagine and MATLAB＇s advantages in mathematics morphology.

关 键 词：MATLAB软件 ALOS遥感影像 震害建筑物 自动识别 数学形态学 都江堰市区 

分 类 号：P315.9[天文地球—地震学] TP75[天文地球—固体地球物理学]