基于COSMO-SkyMed和SPOT-5的城镇洪水淹没信息快速提取研究  

A Research on Town Flood Information Rapid Extraction based on COSMO-SkyMed and SPOT-5

作　　者：王嘉芃 刘婷[1,2] 俞志强[3] 胡潭高[1,2] 张登荣[1,2] 寻丹丹 王冬海 

机构地区：[1]杭州师范大学遥感与地球科学研究院,浙江杭州311121 [2]浙江省城市湿地与区域变化研究重点实验室,浙江杭州311121 [3]浙江省地理信息中心,浙江杭州310012 [4]中电科海洋信息技术研究院有限公司,北京100043

出　　处：《遥感技术与应用》2016年第3期564-571,共8页Remote Sensing Technology and Application

基　　金：浙江省公益技术研究社会发展项目(2014C33049);国家863计划(2013AA12A402;2015AA7015096);杭州市科技发展计划(20150533B04)资助

摘　　要：洪水是我国最为频繁的自然灾害之一,如何快速准确地获取洪水淹没范围在救灾减灾工作中具有重要意义。目前,卫星遥感技术已广泛应用于洪水信息提取的研究中。不同的遥感数据源在洪水信息提取中各有利弊,综合研究雷达影像和可见光影像的优缺点,建立了基于多源遥感数据的洪水淹没信息快速提取模型。首先,利用灾中第一时间获取的COSMO-SkyMed雷达影像,采用面向对象的方法提取出洪灾发生时的水域空间信息;其次,利用灾前SPOT-5高分辨率光学影像,采用多光谱影像波段运算和决策树分类的思想提取出常态下的水域空间信息;最后,对灾中雷达影像COSMOSkyMed提取的水体和灾前光学影像SPOT-5提取的水体进行空间差值运算,得到洪水淹没范围信息,并利用洪水当天拍摄的无人机遥感影像对结果进行精度评价。将该模型应用于2013年浙江余姚水灾,监测结果表明:在洪水发生后,能够快速获取淹没范围空间信息,并且提取精度达到93.7%,为洪灾的防治以及抗洪抢险救灾工作提供强有力的技术支撑和基础数据信息。The flood is one of the most frequent natural disasters in our country,and how to get flooded areas quickly and accurately is of great significance in the disaster relief work.At present,the satellite remote sensing technology has been widely used on the study of floods information extraction.Different remote sensing data sources have their own advantages and disadvantages in floods information extraction,so this paper synthetically studies superiority and inferiority of the radar image and the visible light image,modeling floods information quick extraction based on multi-source remote sensing data.First of all,extract the spatial information of water on the COSMO-SkyMed radar image during floods using the object-oriented method;Secondly,extract the spatial information of normal water on the SPOT-5high resolution optical image before floods combining band math of the multispectral image with decision tree classification;Finally,the flood submerging range is obtained by using differential technology of water spatial information extracted from SPOT-5before floods and COSMO-SkyMed during floods,and then assess accuracy of final results using UAV（Unmanned Aviation Vehicle）remote sensing images taken on the same day of the COSMO-SkyMed radar image.Applied to Yuyao＇s floods in Zhejiang in 2013,this model is able to get floods submerging scope quickly,and the extracting accuracy reaches 93.7%.Furthermore,the model will provide strong technical support and basic data information for flood control and flood fighting and disaster relief work.

关 键 词：COSMO-SkyMed SPOT-5 洪水淹没范围 面向对象 差值运算 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]