Near-field 3-D synthetic aperture radar imaging via compressed sensing

This paper successfully implements compressed sensing (CS) to a near-field wideband 3-D synthetic aperture radar (SAR) imaging system. SAR data are measured at a low percentage of random-selected positions on a uniform grid of planar aperture in the stripmap mode. The near-field 3-D range migration algorithm (RMA) is used in combination with the CS principle to reconstruct the 3-D image via l₁ regularized least-square approach. Experiments were performed with Q-band stepped-frequency monostatic stripmap SAR imaging system on a blue foam embedded with eight rubber pads and one copper square chip. The results of the experiments show near-field 3-D image of the specimen under test (SUT) can be reconstructed efficiently from low percentage of the full measurement positions, which largely lessens the data collection load. The reconstructed image was better focused and denoised.