Abstract:Distributed scatterer interferometry (DSI) is a well-known technique for ground surface deformation monitoring. Central to this process, phase estimation reconstructs a consistent phase series from all interferometric combinations. In theory, the maximum likelihood estimator (MLE) is the optimum approach for phase estimation. However, in practice, its performance is often compromised. Previous studies have demonstrated that the coherence magnitude bias is a source of error. However, other sources of error in the MLE processing remain unclear. This study systematically assesses the sources of error in phase estimation and develops a hybrid approach that corrects three identified sources of error: 1) To address the error from inhomogeneous pixels, an algorithm based on the covariance matrix preestimation and general likelihood ratio test (CMGLR) is developed to select more accurate homogeneous pixels; 2) to mitigate the bias from coherence magnitude matrix, we apply the oracle approximating shrinkage (OAS) algorithm to estimate the precision matrix with higher accuracy; and 3) to tackle the noise from interferometric phase matrix, the filtering principles are defined and the covariance matrix filtering (CMF) algorithm is designed to suppress the noise. A series of simulated experiments demonstrate the effectiveness of the proposed approach. Additionally, a real TanDEM-X experiment shows that the proposed approach can reconstruct the time series phase with reduced noise. Furthermore, the estimated deformation exhibits improvement with significantly increased measurement points MPs (>2.4 times) and higher accuracy compared to the traditional method based on the Kolmogorov–Smirnov (KS) test and sample covariance matrix (SCM). Particularly, it exhibits exceptional performance in monitoring fine structures, while the traditional method usually fails with very few MPs. These results underscore the significant potential of this approach in the realm of ground surface deformation monitoring.

Robust Covariance Matrix Estimation with Application to Volcano Monitoring Using SAR Image Stacks

Robust Estimators for Multipass SAR Interferometry

The robust InSAR optimization framework with application to monitoring cities on volcanoes

A Region-Based Method of Vegetation Coverage Extracting in Complex Terrain Areas Using Polarmetric SAR Data.

Fast Cauchy-Rician Modelling of SAR Images with Method of Algebraic Moments Estimator

Robust Rayleigh Regression Method for SAR Image Processing in Presence of Outliers

Robust Estimation of Measurement Error Covariance

Advanced Coherence Stacking Technique Using High Resolution TerraSAR-X Spotlight Data

A Robust Probabilistic Generative Method for Multitemporal Interferometric Phase Reconstruction: RPGSAR

Time-Series InSAR Dynamic Analysis with Robust Sequential Adjustment

Robust Covariance Matrix Estimation for Radar Space-Time Adaptive Processing (STAP)

Optimal Estimation of Distributed Scatterer Phase History Parameters from Meter-Resolution SAR Data.

A statistically homogeneous pixel selection approach for adaptive estimation of multitemporal InSAR covariance matrix

Low Rank Approximation Approach for Microwave Radar Coincidence Imaging

An M-Estimation Method for InSAR Nonlinear Deformation Modeling and Inversion

A Matrix Completion Based Method for RFI Source Localization in Microwave Interferometric Radiometry

Model Quality Aware RANSAC: A Robust Camera Motion Estimator

A Hybrid Approach for High-Precision Phase Estimation in Distributed Scatterer Interferometry

Markovian models for one dimensional structure estimation on heavily noisy imagery

Estimation of Ground Deformation of Transmission Corridors in Mountain Areas Using Stanford Method.

Online High Resolution Stochastic Radiation Radar Imaging Using Sparse Covariance Fitting