Multiframe Phase-Guided Dual-Layer Superresolution Imaging Framework for Scanning Radar
Abstract
Scanning radar employs antenna scanning to facilitate ground observation and offers real-time observation advantages for disaster early warning and building mapping. Conventional angular superresolution imaging techniques typically process each frame independently to enhance surveillance capability. However, these methods are vulnerable to the effects of background clutter and moving targets, reducing the mapping performance of buildings. To address these challenges, this article proposes a multiframe phase-guided dual-layer superresolution imaging framework for scanning radar. First, a multiframe phase-guided dual-layer strategy is proposed to effectively distinguish stable scatterers from dynamic scatterers, including background clutter and moving targets. It leverages the phase-preserving characteristics of stable scatterers across multiple frames to facilitate their differentiation from the background. Second, the stable-layered scattering echoes are coherently accumulated to improve image quality by enhancing the echo signal-to-noise ratio. Finally, based on the multiframe stable-layered echo data, this article presents the missing data sparse covariance-based estimation method to address incomplete echoes, which can offer a highly efficient and robust solution by decomposing the covariance matrix, which facilitates rapid inversion of the matrix through the Woodbury matrix identity and conjugate gradient techniques. Simulation and experimental results demonstrate that the proposed method effectively reduces the impact of background clutter and moving targets on superresolution imaging performance.