摘要
针对重复轨合成孔径雷达差分干涉测量中的失相干和大气扰动问题,介绍一种基于短基线差分干涉纹图集的地表形变时间序列及形变速率提取算法。选择短基线干涉像对构成差分干涉纹图集,以均值相干系数阈值为长时间间隔下相干目标识别指标,将离散相干点目标构建成Delaunay三角网,利用最小费用流算法完成相位解缠;视大气扰动引起的相位延迟为随机误差,以奇异值分解法对形变相位序列进行最小二乘处理解算形变累积量,通过最小二乘回归得到相干目标的形变速率。试验结果证明该算法对长时间序列形变场时空变化连续监测有效。
Temporal decorrelation and phase delay are two main abstacles for repeat pass differential SAR interferometry (D-InSAR) used for surface deformation monitoring. We presented an algorithm for deformation timeseries retrieval and linear-rate estimation by use of short baseline differential interferograms stack. The pixel candidates presenting a good coherence level in the whole set of interferograms stack were identified and the resulting nonuniform meshes were tessellated with the Delauney triangulation to establish connections among them. The Minimum Cost Flow(MCF) algorithm was used for phase unwrapping, and the Singular Value Decompostion (SVD) method was applied to link the indenpendent interferograms subset, thus obtained the deformation time-series of each coherent pixel. The linear deformation rate was calculated from the time series with a least square regression. Presented results achieved on ENVISAT SAR data confirms the validity of the proposed approach for long-term deformation evlution monitoring.
出处
《大地测量与地球动力学》
CSCD
北大核心
2008年第2期61-66,共6页
Journal of Geodesy and Geodynamics
基金
国家高技术"863"计划项目(2007AA12Z171)
中国地质调查局计划项目(1212010560705
1212010540905)
欧空局CAT-1项目(ProjectID:C1P.3863)
关键词
D—InSAR
失相干
大气扰动
奇异值分解
干涉纹图集
D-InSAR
decorrelation
atmosphere disturbance
singular value decomposition
interferogram stacks