NL-MMSE: A Hybrid Phase Optimization Method in Multimaster Interferogram Stack for DS-InSAR Applications

Abstract

When the distributed scatterer interferometric synthetic aperture radar (DS-InSAR) technology is used for surface deformation monitoring, the accuracy strongly depends on the quality of phase optimization. Especially in the low coherence region, how to conveniently and effectively improve the quality of phase optimization has been a hot and difficult research topic in recent years. This paper proposes a hybrid phase optimization method for the DSInSAR technology, which chooses either non-local (NL) or minimum mean square error (MMSE) method for each pixel according to the distribution of statistically homogeneous pixel (SHP) in two windows of different sizes. This hybrid method (NL-MMSE) is not limited by the number of interferogram and is not influenced by multi-master images. The NL-MMSE method was applied to the deformation monitoring in the Jinsha River basin, Tibet, China, using 28 Sentinel-1A SAR images acquired between February and December 2020. Compared with the NL and MMSE methods, the NL-MMSE method provides better phase quality of the interferogram stack and is able to extract more temporal coherence points for subsequent deformation inversion. The deformation monitoring results showed there are three obvious large-scale landslide deformation and dozens of small-scale deformation in the study area. It is demonstrated that the NL-MMSE method based on DSInSAR technology can accurately monitor the detailed deformation characteristics of the low coherence surface, and can be applied and promoted as an effective means of identifying and monitoring geological hazards.

Identification and Analysis of Deformation Areas in the Construction Stage of Pumped Storage Power Station Using GB-InSAR Technology

“…2) shows that the pixels excluded by the intensity deviation threshold are basically located in regions without backscattering or noise; the SHPs obtained by FaSHPS does not distinguish well between noisy and strongly scattering regions and are worse than the BWS test; the SHPs obtained by D I + FaSHPS compensate well for the shortcomings of FaSHPS and do not differ significantly from D I + BWS. Based on the number of SHPs selected, the pixels can be divided into PSs and distributed scatterers [44], [45]. Previous studies and analyses have concluded that the maximum number of SHPs considered as PSs is set to 10 [37], [40].…”

Inversion of Reservoir Parameters for Oil Extraction Based on Deformation Monitoring With InSAR

“…2). Conventional time-series InSAR techniques are limited to obtain fine monitoring results in target areas, while DS-InSAR can increase the spatial density of the measurement points with a large number of randomly distributed scatters to improve the sampling rate effectively [47][48]. It is better suited for specific characteristics of our study area.…”

An Automatic Method for Delimiting Deformation Area in InSAR Based on HNSW-DBSCAN Clustering Algorithm

“…Due to the high vegetation coverage in the study area, the traditional PS point selection method is used to achieve better results. This paper uses three-point selection strategies to select PS points to obtain target points with high phase quality and comprehensive coverage, namely, amplitude departure index method, spectral coherence, and phase stability [26,27].…”