Hidden target reconstruction is a novel imaging technique applicable to the fields of autonomous driving, urban street combat, and urban anti-terrorism. In this paper, we propose a non-line-of-sight (NLOS) inverse synthetic aperture radar (ISAR) imaging algorithm dubbed NLOS moving target compressed sensing (NMTCS) for the moving targets based on Bayesian theory by exploiting the multipath effect and the scattering properties of NLOS scene. For difficult-to-extract hidden target echoes, the high-pass filter is used to suppress the clutter. For sparse echoes that are difficult to reconstruct, a compressed sensing (CS) algorithm that exploits the scattering properties of NLOS echoes is proposed. Mirror symmetry is used to recover the real target. Finally, millimeter-wave (MMW) measured data and traditional algorithms are employed to verify the effectiveness and superiority of NMTCS.