Low rank plus sparse decomposition of synthetic aperture radar data for maritime surveillance

Abstract

Synthetic Aperture Radar (SAR) systems produce a tremendous amount of redundant data if persistent radar surveillance of a specific area is implemented. This paper performs an efficient data reduction extrapolating maritime targets in motion from background subtraction. The technique is based on Robust Principal Component Analysis (RPCA). The algorithm is implemented by Convex Programming (CP). This Low Rank and Sparse Decomposition (LRSD) activity permits the separation of sparse objects of interest, with a stationary low-rank background. RPCA applied to SAR surveillance permits the saving of a large amount of data. Dynamic SAR is procured by Multi Chromatic Analysis (MCA) of Native (RAW)1 satellite data.