Phase unwrapping (PU) is a key program in data processing in the interferometric synthetic aperture radar (InSAR) technique, and its accuracy directly affects the quality of final SAR data products. However, PU in regions with large gradient changes and high noise has always been a difficult problem. To overcome the limitation, this article proposes an adaptive square-root unscented Kalman filter PU method. Specifically, a modified phase gradient estimation (PGE) algorithm is proposed, in which a Butterworth low-pass filter is embedded, and the PGE window can be adaptively adjusted according to phase root-mean-square errors of pixels. Furthermore, the outliers of the PGE results are detected and revised to obtain high-precision vertical and horizontal phase gradients. Finally, the unwrapped phase is calculated by the adaptive square-root unscented Kalman filter method. To the best of our knowledge, this article is the first to combine the modified PGE with an adaptive square-root unscented Kalman filter for PU. Two sets of simulated data and a set of TerraSAR-X/TanDEM-X real data were used for experimental verification. The experimental results demonstrated that the various improvement measures proposed in this article were effective. Additionally, compared with the minimum-cost flow algorithm (MCF), statistical-cost network-flow algorithm (SNAPHU) and unscented Kalman filter PU (UKFPU), the proposed method had better accuracy and model robustness.
20 April 2023 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Phase unwrapping (PU) as an ill-posed inverse scientific problem has invariably been the significant difficulty in interferometric synthetic aperture radar (InSAR) technique for topographic mapping. Conventional PU approaches may produce a local or global unwrapping bias in severe phase changes caused by noise and large gradient terrain because the phase continuity assumption is not satisfied. To address this issue, we have created a novel phase quality-guided integrated filtering and unwrapping process for PU. In particular, the PU path is guided by an innovative phase cost indicator, which can automatically coordinate contributions from several phase quality estimation methods so that the unwrapping path can circumvent phase discontinuity regions and delay unwrapping these pixels. Since the central difference transform is another extra effective nonlinear transformation mode based on sigma points and only has one regulation parameter, we have employed the central difference information filter (CDIF) for the first time to realize the dynamic estimation of the unwrapped phase in the PU process. We tested the proposed approach using two real datasets with different mountainous areas. The results demonstrate that the PU accuracy of the proposed approach based on these two datasets is improved by 25.2% and 40.4% compared to the unscented information filtering (UIF) PU method, and the anti-noise capacity and unwrapping efficiency of the CDIF are both slightly superior to the UIF.
14 August 2023 in IEEE Journal on Miniaturization for Air and Space Systems
Phase unwrapping (PhU) of the large gradient deformation in the coal mining subsidence center has always been the main problem in the differential interferometric synthetic aperture radar (DInSAR) data processing. The accuracy of PhU will directly affect the deformation results of the mining subsidence center. To address this issue, in this article, we proposed a two-step PhU method which combines $L^{1}$ -norm and $L^{2}$ -norm. This method can effectively obtain the PhU results of the large gradient deformation in the mining subsidence center. First, the filtered DInSAR interferometric phase is unwrapped using $L^{2}$ -norm, and the first-step unwrapped phase is obtained. Then, the first-step unwrapped phase is rewrapped, which performs conjugate multiplication with the original interferometric phase to obtain the residual phase. Moreover, the residual phase is unwrapped by the $L^{1}$ -norm method to obtain the second-step unwrapped phase. Finally, the final unwrapped phase is obtained by summing the first-step and second-step unwrapping results. Experiments are conducted with simulated and GaoFen-3 SAR data sets. To compare against the representative PhU method, the proposed method can effectively solve the problem of PhU in the large gradient deformation of the mining areas.
20 November 2025 in International Journal of Applied Earth Observation and Geoinformation
Interferometric synthetic aperture radar (InSAR) is the preferred technique for monitoring three-dimensional (3D) surface displacements. However, conventional InSAR phase unwrapping (PhU) methods face challenges in interpreting local low-coherence regions and dense fringe patterns, reducing the reliability of subsequent 3D decomposition. Additionally, the North-South (N-S) displacement retrieved from InSAR measurements is often inaccurate due to the near-polar orbits of radar satellites. To address these issues, this paper presents a novel method for retrieving mining-induced 3D displacements from dual InSAR geometries. First, we propose a two-step PhU approach based on the central difference information filter phase unwrapping (CDIFPU), which can generate a complete deformation field. Subsequently, vertical and East-West (E-W) displacements are separated using InSAR measurements from dual imaging geometries. Finally, we use an extensible subsidence facet kernel to retrieve the N-S displacement based on random medium theory. Simulated experiments show that the two-step PhU improves unwrapping accuracy by approximately 78.4%, 64.8%, and 59.5% over the branch-cut, MCF, and CDIFPU methods. The accuracy of the two-step PhU in retrieving vertical and east-west displacements is 65.8% and 71.3% higher than that of the MCF. Moreover, the subsidence facet kernel further improves the N-S displacement accuracy by 68.5% compared to the conventional first-order difference operators. Experiments with RADARSAT-2 and TerraSAR-X images from the Fengfeng Mining Area, China, demonstrate that the two-step PhU is more effective than SNAPHU, PUMA, and one-step deep learning PhU in retrieving absolute phase in local low-coherence regions, enhancing the quality of vertical and E-W displacements. Ultimately, its N-S displacement is successfully extracted using the subsidence facet kernel.