Backprojection Imaging of Moving Objects

Abstract

A backprojection-based synthetic aperture radar (SAR) processing approach for fast and accurate imaging of moving objects and stationary background is proposed. This divide-and-conquer backprojection approach allows for efficient detection of moving objects, estimation of their positions and velocities, and overlaying their focused images over the stationary background. Two algorithms are presented to exploit the tradeoff between the complexity and performance. The full 4-D phase space imaging algorithm provides the optimal performance under low signal-to-clutter conditions, while the adaptive detection and imaging leads to reduced complexity in more favorable scenarios. We show that in a single-channel SAR the conventional straight-line data collection path does not resolve the target's velocity, therefore a curved path is assumed in the formulation and numerical examples.