Matrix-Completion-Based Airborne Tomographic SAR Inversion Under Missing Data

Tomographic synthetic aperture radar imaging (TomoSAR) is a new SAR imaging modality that extends the synthetic aperture principle into the elevation direction. In TomoSAR, the elevation resolution depends on the length of elevation aperture and the baseline distribution. When the length of elevation aperture is fixed, the number of baselines is important for the elevation reconstruction. Therefore, improving the elevation imaging quality with the finite amount of baselines is worth researching. This letter proposes a novel missing data compensation approach in TomoSAR via matrix completion (MC), which is a technique of the low-rank matrix recovery from a subset of the matrix elements. In the proposed method, we first exploit MC to estimate the 2-D focused image data of unknown baselines, which are located on the uniform data grid without changing the length of elevation aperture. Based on the recovered data, we then recover the elevation reflectivity function by the spectral analysis (SA) method. Compared with the conventional SA technique, the proposed approach can achieve much higher elevation image quality, due to more available data offered by MC. The effectiveness of the proposed method is validated with the experimental results based on simulated and real airborne data.