CMTI: Non-Line-of-Sight Radar Imaging for Non-Cooperative Corner Motion Target

Concealed moving target imaging is an important task in the fields of autonomous driving, urban combat, city anti-terrorism, etc. However, existing radar imaging techniques mainly focus on line-of-sight (LOS) targets and struggle to sense non-line-of-sight (NLOS) moving targets. In this paper, a joint NLOS model and inverse synthetic aperture radar (ISAR) algorithm, dubbed corner motion target imaging (CMTI), is proposed for moving target imaging via millimeter-wave (MMW) radar. In this scheme, initially, to separate hidden motion target echoes and background clutter, we utilize a low-frequency filtering method to suppress the clutter and an NLOS echo extraction method based on OTSU to extract the target echoes. Subsequently, to solve the problem of focusing and imaging difficulties, a novel NLOS imaging method based on the polar format algorithm (PFA) and the prominent point processing (PPP) autofocus algorithm is exploited. The method consists of a cross-correlation method and a minimum variance algorithm for envelope alignment and phase adjustment, the minimization entropy autofocusing (MEA) method and PPP for residual phase error and target rotational speed estimation, and PFA for 2-D high-resolution imaging. Eventually, to obtain the real target image, we present a reconstruction algorithm based on the mirror symmetry criterion. Near-field experiments are set up to verify CMTI. The experimental results demonstrate that CMTI can provide high-resolution images of the NLOS moving targets.