Sparse Millimeter-Wave InSAR Imaging Approach Based on MC

Millimeter-wave interferometric synthetic aperture radiometer (InSAR) imaging has received considerable attention owing to its pure radiation receiving scheme and high resolution. In recent years, compressive sensing (CS) has been applied to InSAR on the assumption that InSAR image could be formulated to a sparse vector. However, the CS-based InSAR ignores the 2-D sparsity of images and its performance degrades while processing large-scale images. Further using the sparsity of InSAR image, a novel InSAR imaging approach referred to as InSAR-matrix completion (MC) is proposed in this letter by adopting the MC techniques, which can recover a low-rank matrix from a small subset of its corrupted entries. Directly representing the target image as a matrix, InSAR-MC formulates the estimation of the visibility function as a 2-D linear transformation and inversely obtains the target image by convex optimization as sparsity of millimeter-wave image and interferometric operation satisfy the low-rank property. Experimental results demonstrate the proposed approach's validity and better performance of the recovered images with comparison against CS-based InSAR and other traditional approaches.