Near-field Radar Imaging Acceleration Strategy Using Wavelet Transform and Multiscale Processing

With the continuous advancement of millimeter-wave (mmWave) radar technology, high-resolution, high-definition and real-time near-field imaging has emerged as a new research objective. However, existing near-field SAR imaging approaches still face numerous challenges. Achieving high-resolution imaging requires optimizing and enhancing radar bandwidth, data acquisition density, and scanning aperture. This, in turn, imposes higher demands on processing speed and parallelism for computational units to enable real-time imaging capabilities. In this paper, we propose a novel multiscale decomposition imaging acceleration strategy based on the wavelet transform, which achieves parallel acceleration across multiple units while reducing the computational cost. By comparing imaging results between this method and previous algorithms using metrics such as SSIM and Jaccard, we demonstrate high-precision, rapid radar imaging of two-dimensional planes with minimal compromise in imaging quality.