RSOFNet: PolSAR Target Recognition Using a Dual-Branch Framework Based on Rotational Similarity Features and Span-Conditioned Modulation

Abstract

Restricted by view sensitivity and limited data volume, current polarimetric SAR vehicle target recognition still faces bottlenecks, including inadequate characterization of scattering properties and poor generalization performance. To address these limitations, this article proposes a rotation-domain similarity feature (RSF) to precisely characterize the angular dependence of target scattering characteristics and introduces an RSF-span overview-focusing network. The RSF matches the target coherency matrix with canonical scattering templates across the continuous rotation domain, yielding angle-dependent responses with explicit physical meaning. A span branch extracts sample-level scene context from the total-power image to guide RSF feature learning in a sample-adaptive top-down manner. A physics-constrained fusion layer integrates RSF-derived physical statistical priors, span-modulated mechanism-discriminative weights, and context-conditioned template reliability weights, ensuring physically interpretable feature embeddings. Experiments on the High-Resolution Airborne PolSAR Vehicle Target Recognition and GOTCHA datasets indicate that the proposed method achieves performance superior to data-driven, single-feature, and multifeature fusion baselines under standard, cross-depression-angle, and reduced-training-data conditions, with the advantage becoming more pronounced as sample size decreases.