Skip to content
Open access

Radon-Root-MUSIC: A Super-Resolution Method for PolSAR Ship Detection in Dense Target Scenarios

2026 · IEEE Transactions on Aerospace and Electronic Systems · Vol 62, pp. 18055-18073 · 0 citations · 34 references

Abstract

In polarimetric synthetic aperture radar (PolSAR) image, ship detection is inherently constrained by the Rayleigh resolution limit imposed by finite system bandwidth. This limitation restricts the accurate discrimination of closely spaced targets, thereby leading to missed detections, false alarms, and target shape distortions. To overcome these challenges, this article proposes a super-resolution method for PolSAR ship detection in dense target scenarios by integrating the Radon transform with the Root-MUSIC algorithm. First, a coarse target region selection is performed to identify the approximate target area, and an angular strip extraction strategy is employed to extract data strips aligned with each projection direction, thereby reducing noise interference and computational complexity. The Radon transform is then applied to project the 2-D PolSAR images from the four polarimetric channels into 1-D signals along the range, azimuth, main diagonal, and secondary diagonal directions. Appropriate projection angles are selected along the diagonal directions to avoid nonequidistant sampling. The resulting 1-D signals from the four polarimetric channels are transformed to the Fourier frequency domain and used to construct covariance matrices. The Root-MUSIC algorithm is then employed to estimate target projection positions in each direction based on the constructed covariance matrices, with an optimized strategy for noise subspace construction and projection position quantity estimation incorporated to enhance estimation accuracy. Finally, a four-line intersection positioning strategy is adopted to integrate the estimated positions from the four directions, thereby achieving accurate detection of ship targets. The effectiveness of the proposed method is verified through simulated data and real airborne PolSAR data, as well as PolSAR data from GF-3 and RadarSAT-2.

Read PDF

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.