利用光学遥感影像光流场模型进行地表形变分析  

Analysis of Surface Deformations on the Basis of Optical Flow Field Models from Optical Remote Sensing Images

作　　者：丁明涛[1,2,3,4,5] 陈浩杰 李振洪 刘振江 DING Mingtao;CHEN Haojie;LI Zhenhong;LIU Zhenjiang(College of Geological Engineering and Geomatics,Chang'an University,Xi'an 710054,China;Key Laboratory of Loess,Xi'an 710054,China;Big Data Center for Geosciences and Satellites,Xi'an 710054,China;Key Laboratory of Ecological Geology and Disaster Prevention,Ministry of Natural Resources,Xi'an 710054,China;Laboratory of Smart Earth,Beijing 100029,China)

机构地区：[1]长安大学地质工程与测绘学院,陕西西安710054 [2]黄土科学全国重点实验室,陕西西安710054 [3]长安大学地学与卫星大数据研究中心,陕西西安710054 [4]自然资源部生态地质与灾害防控重点实验室,陕西西安710054 [5]智慧地球重点实验室,北京100029

出　　处：《武汉大学学报（信息科学版）》2024年第8期1314-1329,共16页Geomatics and Information Science of Wuhan University

基　　金：国家重点研发计划(2021YFC300400);国家自然科学基金(42374027);陕西省科技创新团队(2021TD-51);陕西省地学大数据与地质灾害防治创新团队(2022);智慧地球重点实验室基金(KF2023YB04-01);高分专项川藏区域综合治理应用与规模化产业化示范项目(87-Y50G28-9001-22/23);中央高校基本科研业务费专项资金(300102262203,300102262902)。

摘　　要：光学遥感影像像素偏移量追踪是反演同震形变场和监测滑坡的一种重要手段。基于相关性匹配的传统像素偏移量追踪方法,通过搜索相关性最强的匹配窗口估计中心像素的位移,计算效率低且在大梯度形变区域失相关现象严重,存在形变边界提取不精确的问题。为高效获取精确的地表形变,将计算机视觉领域的光流场模型引入像素偏移量追踪问题,提出适用于光学遥感影像反演地表形变的光流场方法,给出像素偏移量时序反演的加权改进算法。通过塔吉克斯坦同震形变场模拟实验,评估光流场方法估计地表形变场的可行性及其在最小可探测形变方面的性能;通过加州地震同震形变场反演和白格滑坡偏移量估计实验,讨论光流场方法的计算效率优势和形变区域提取的精确性;通过白格滑坡时序形变分析,进一步论述利用光流场方法估计大梯度形变的有效性和时序反演加权改进算法的鲁棒性。结果显示,相比于传统窗口相关性匹配方法,光流场方法的偏移量追踪精度为0.032像素,计算效率提升了20倍左右,形变区范围提取精度提升了25.9%;改进的加权时序反演算法将光学遥感影像东西向和南北向位移估计的不确定性分别降低了16.2%和12.4%。Objectives:Pixel offset tracking(POT)for optical remote sensing imagery is widely used to invert coseismic deformation fields and monitor landslides.Traditional pixel offset tracking method estimates the displacement of the central pixel by searching for the matching window with the highest correlation,which is computationally inefficient and suffers from inaccurate deformation boundary extraction due to the decoherence effects in the region with dynamical deformation. We introduce the optical flow field modelcommonly used in computer vision to the pixel offset tracking problem to obtain accurate surface deformationefficiently. Methods: The optical flow field method applicable to optical remote sensing images and theimproved inversion algorithm for the time series analysis are proposed to inverse the surface deformation.Experiments on the simulated coseismic deformation fields in Tajikistan are detailed to assess the feasibilityand the minimum detectable deformation of the optical flow field method. The advantages of the proposedmethod over computational cost and deformation boundary extraction accuracy are illustrated by the co-seismicdeformation field of the California earthquake and the displacement of the Baige landslide. Furthermore,the performance on estimating large gradient deformation and the robustness of the improved time seriesinversion algorithm are discussed by analyzing the time series deformation of the Baige landslide. Re⁃sults: The results show that compared with the traditional window correlation matching method, the opticalflow field method has an offset tracking accuracy of 0.032 pixel, which improves the computational efficiencyby about 20 times, and the accuracy of the deformation zone is improved by 25.9%. The time seriesweighted inversion algorithm reduces the uncertainties in the estimation of east-west and north-south displacementsof optical remote sensing images by 16.2% and 12.4%, respectively. Conclusions: The proposedmethod alleviates the pixel offset tracking problem in the bound

关 键 词：光流场方法 滑坡 同震形变 计算效率 加权时序反演算法 

分 类 号：P237[天文地球—摄影测量与遥感]