Wall artifacts removal for target imaging enhancement in UWB through-the-wall radar application

摘要

In ultra-wideband through-wall-imaging applications, wall artifacts are always much stronger than the target reflections, and they tend to persist over a long duration of time. Therefore, targets are obscured and not visible in the image. The existing wall artifacts removal methods such as the moving average subtraction method, spatial filtering method and singular-differential decomposition method can efficiently remove wall artifacts in synthetic aperture radar (SAR) mode. However, with multi-input-multi-output (MIMO) radar widely used, the signal model and corresponding characteristics of wall artifacts are significantly different from those in the SAR mode. As a result, the existing methods are seldom adopted because their performances are severely affected. To address this problem, this paper proposes a novel wall artifacts removal method that is based on singular value decomposition and a cost-function 'varimax norm'. The proposed method is tested with synthetic and experimental data in monostatic SAR mode and MIMO mode. Compared with existing methods using the signal-to-clutter ratio of the processing results, it is demonstrated that the proposed method can efficiently remove wall artifacts in both modes. Particularly in MIMO mode, the proposed method is much more capable than the existing methods in mitigating artifacts and enhancing target imaging.