摘要
滑坡是我国最严重的地质灾害之一。针对常规MTInSAR技术在开展滑坡形变监测时面临的监测点密度不足、监测精度受限及滑坡解译受阻的问题,提出一种有效融合PS及DS点的PS/DSInSAR技术。首先,开展HTCI同质像素识别及相干性n次幂加权的相位优化处理,并筛选DS候选点;然后,将其与振幅离差获取的PS点进行融合并筛选高质量监测点,通过对监测点的相位解译提取滑坡形变信息。以云南省德钦县为研究区开展相关滑坡监测分析。实验结果表明,PS/DSInSAR技术获取的监测点密度较常规的StaMPS-SBAS技术提升了约14倍,可以更精准地识别滑坡边界及开展时序演化分析,且分析表明,研究区滑坡与季节性降雨具有显著的相关性。
Landslide is one of the most serious geological disasters in China.In response to the problems the conventional MTInSAR technique faces in monitoring landslide deformation,such as insufficient monitoring point density,limited monitoring accuracy,and difficulty in landslide interpretation,a PS/DSInSAR technique that effectively integrates PS and DS points is proposed.Firstly,the proposed technique carries out HTCI homogeneous pixel recognition and phase optimization processing with coherence n-power weighting and obtains DS candidate points.Then,it fuses them with PS points obtained from amplitude deviation and screens high-quality monitoring points.Finally,the landslide deformation information is extracted through phase interpretation of monitoring points.Taking Deqin county,Yunnan province,as the study area,relevant landslide monitoring and analysis are conducted.The experimental results show that the density of monitoring points obtained by the PS/DSInSAR technique increases by about 14 times compared to the conventional StaMPS-SBAS technique,which can more accurately identify landslide boundaries and conduct temporal evolution analysis.And analysis shows a significant correlation between landslides and seasonal rainfall in the study area.
作者
周仿荣
马朋序
文刚
高瑞
ZHOU Fangrong;MA Pengxu;WEN Gang;GAO Rui(Electric Power Research Institute of Yunnan Power Grid Co.Ltd.,Kunming 650217,China;Jinan Geotechnical Investigation and Surveying Institute,Jinan 250101,China;Shandong Academy of Agricultural Sciences,Jinan 250100,China)
出处
《遥感信息》
CSCD
北大核心
2024年第1期43-51,共9页
Remote Sensing Information
基金
中国南方电网云南电网有限责任公司科技项目(YNKJXM20220151)。
作者简介
周仿荣(1982—),男,教授级高级工程师,主要研究方向为卫星遥感电力应用、雷达图像处理及干涉测量。E-mail:42783590@qq.com;通信作者:马朋序(1995—),男,硕士研究生,主要研究方向为InSAR时序形变监测、测绘数据处理。E-mail:pengxuma@126.com。
