Maritime Radar Target Multifeature Detection Optimization Method Under Capability Lower Bound Constraint

Multi-feature detection algorithm has demonstrated promising performance in low-RCS radar target detection. However, the multi-feature detection method without capability lower bound constraint usually leads to uncertainty in detection result and inefficient allocation of sensing resource in practical applications. To cope with this problem, a radar target multi-feature detection optimization method under capability lower bound constraint is proposed to enhance the robustness and real-time performance of maritime radar target detection in this paper. Without loss of generality, signal-clutter-noise ratio (SCNR) is employed as a representative metric to characterize the distinction between target and clutter, serving as the proxy for the capability lower bound. Firstly, a surrogate model is established to map the relationship between SCNR and detection results by means of the measured radar data. Secondly, an improved average tail margin (ATM) algorithm combined with feature importance analysis is applied to accurately estimate the lower bound, which is adopted to estimate an adaptive detection threshold for suppressing a large amount of background clutter. Finally, the multi-feature detection algorithm finely discriminates target from background clutter. Extensive experiments are carried out both public radar dataset provided by Radar Journal and our measurement radar dataset captured at outfield. The experimental results show that the proposed method not only reduces computational complex with superior real-time detection performance, but also can enhance the robustness without sacrificing detection accuracy, outperforming baseline multi-feature methods including conventional CFAR. This work enhances the reliability and the practical applicability of sea-surface radar target detection, and it demonstrates strong generalization ability under complex maritime environment.