基于特征点的SAR影像偏移跟踪方法研究  

Research on SAR Image Offset Tracking Method Based on Feature Points

作　　者：彭林才 王华[1] 吴希文 Peng Lincai;Wang Hua;Wu Xiwen(School of Civil and Transportation Engineering,Guangdong University of Technology,Guangzhou 510006,China)

机构地区：[1]广东工业大学测绘工程系,广东广州510006

出　　处：《城市勘测》2021年第1期88-93,共6页Urban Geotechnical Investigation & Surveying

基　　金：国家自然科学基金青年基金项目(41104016)《基于InSAR/GPS的青藏高原地壳运动速度场及其变形模型研究》。

摘　　要：在地球科学应用中,合成孔径雷达(synthetic aperture radar,SAR)影像的偏移跟踪在测量大梯度地表形变中起着重要的作用。但是,标准偏移跟踪是估计均匀分布于两个SAR影像上的同名点的变形,不考虑地表的散射特性,在实际的应用中,往往会出现选取的同名点存在粗差、相关性低等情况。本文提出了一种结合了Speeded Up Robust Feature(SURF)和偏移跟踪的方法。我们首先使用SURF从SAR影像中探测特征点,然后执行偏移跟踪以获得像素偏移。模拟实验以及尼泊尔地震实验结果表明,该方法能显著提高偏移跟踪的效率、准确性和可靠性。Offset tracking of synthetic aperture radar(SAR)images plays an important role in measuring large gradient surface deformation in geoscience applications.However,the standard offset tracking procedures estimate deformation of tie points uniformly distributed over two SAR images,regardless of the scattering characteristics of the ground surface.In practical applications,there will often be gross errors and low cross-correlationin the selected tie points.In this paper,we propose a method combining Speeded Up Robust Feature(SURF)and offset tracking.We first use SURF to detect feature points from SAR image,and then perform offset tracking to obtain pixel offsets.The results of simulation and Nepal earthquake experiments show that this method can significantly improve the efficiency,accuracy and reliability of the offset tracking.

关 键 词：偏移跟踪 SURF 特征点 配准 

分 类 号：P225.7[天文地球—大地测量学与测量工程] P237[天文地球—测绘科学与技术]