Millimeter-Wave Sparse Imaging for Concealed Objects Based on Sparse Range Migration Algorithm

To relieve the system cost in densely sampling when applying the millimeter-wave (MMW) synthetic aperture radar (SAR) technique for imaging applications, an MMW sparse imaging method is presented by the combination of the under-sampling scheme in space and the following sparse imaging algorithm. Different from the requirement on sub-wavelength sampling interval by Nyquist law, the system only needs to collect a random subset of spatial samples. To transform the spatial samples into wavenumber domain, a sparse range migration algorithm (SRMA) is proposed to realize MMW sparse imaging by embedding the matrix completion (MC) technique into classical RMA. In light of the relationship of Fourier transform between the scattering coefficients and the wavenumber-domain echo, the proposed algorithm utilizes the atomic norm minimization to fulfill the MC from the small set of wavenumber-domain entries to their full data matrix. The experiments are conducted by a wideband MMW transceiver mounted on a linear trajectory for imaging two specimens with different kinds of cover. Their clear imaging contrast of azimuth-range profiles is obtained through SRMA under different under-sampling conditions and RMA with fully sampled data, and verify the effectiveness of the proposed method.