A Sparse-Model-Driven Network for Efficient and High-Accuracy InSAR Phase Filtering

Phase filtering is a vital step for interferometric synthetic aperture radar (InSAR) terrain elevation measurements. Existing phase filtering methods can be divided into two categories: traditional model-based and deep learning (DL)-based. Previous studies have shown that DL-based methods are frequently superior to traditional ones. However, most of the existing DL-based methods are purely data-driven and neglect the filtering model, so that they often need to use a large-scale complex architecture to fit the huge training sets. The issue brings a challenge to improve the accuracy of interferometric phase filtering without sacrificing speed. Therefore, we propose a sparse-model-driven network (SMD-Net) for efficient and high-accuracy InSAR phase filtering by unrolling the sparse regularization (SR) algorithm to solve the filtering model into a network. Unlike the existing DL-based filtering methods, the SMD-Net models the physical process of filtering in the network and contains fewer layers and parameters. It is thus expected to ensure the accuracy of the filtering without sacrificing speed. In addition, unlike the traditional SR algorithm setting the spare transform by handcrafting, a convolutional neural network (CNN) module was established to adaptively learn such a transform, which significantly improved the filtering performance. Extensive experimental results on the simulated and measured data demonstrated that the proposed method outperformed several advanced InSAR phase filtering methods in both accuracy and speed. In addition, to verify the filtering performance of the proposed method under small training samples, the training samples were reduced to 10%. The results show that the performance of the proposed method was comparable on the simulated data and superior on the real data compared with another DL-based method, which demonstrates that our method is not constrained by the requirement of a huge number of training samples.

Automated summary

Phase filtering is a crucial step in InSAR terrain elevation measurements. Existing methods can be divided into traditional model-based and deep learning-based approaches, with deep learning methods often outperforming traditional ones. However, most existing deep learning methods rely heavily on complex architectures and large-scale training sets. In this study, we propose a sparse-model-driven network (SMD-Net) that models the physical filtering process in the network with fewer layers and parameters. The SMD-Net incorporates a convolutional neural network module to adaptively learn the sparse transform, significantly improving filtering performance. Experimental results on simulated and measured data demonstrate that the proposed method surpasses advanced InSAR phase filtering methods in terms of accuracy and speed. Furthermore, even with a small number of training samples, the proposed method still performs comparably on simulated data and outperforms another deep learning-based method on real data, indicating that it is not constrained by the requirement for a large number of training samples.