Spatiotemporal Staggered Projection Imaging for Video-SAR Under the Resource-Constrained Platform

This article focuses on a reliable and high-resolution synthetic aperture radar (SAR) imaging algorithm for lightweight autonomous aerial vehicle (AAV)-borne video-SAR. By combining a spatiotemporal staggered sampling technique and a multiframe enhancement reconstruction network, the proposed spatiotemporal staggered projection algorithm can solve the contradiction between low data transmission rate of lightweight AAV and high frame rate requirement. Specifically, by utilizing the spatial and temporal relationship between multiframe SAR images, the spatiotemporal staggered sampling technique can reduce the system requirements on computational cost and data transmission rate. Furthermore, enhancement reconstruction networks are utilized on the staggered sampled video-SAR frame sequence to fuse a high-resolution SAR image. Real W-band video-SAR experiments are applied to verify the proposed spatiotemporal staggered projection algorithm. U-net and Unet++ are validated to improve the spatiotemporal staggered sampling frame-sequence enhancement reconstruction results, which outperform the spatiotemporal full sampling SAR images in terms of peak signal-to-noise ratio, structural similarity index measure, and equivalent number of looks. From the aspects of computational cost, the proposed imaging algorithm is four or nine times faster than the standard video-SAR imaging algorithm with less quality loss, which is more satisfactory for lightweight AAV platforms.