Offshore wind turbines detection from single-scene SAR imagery using morphological and spatial features

The rapid global expansion of offshore wind power has created an urgent demand for efficient methods to map the deployment scale of offshore wind turbines (OWTs). Synthetic Aperture Radar (SAR) is well suited to this task due to its wide spatial coverage and frequent revisit capability. However, most existing SAR-based detection methods rely on multi-temporal image stacking to suppress dynamic vessel interference or data-intensive deep learning workflows requiring substantial annotated training data and cloud-computing resources. These dependencies introduce update latencies ranging from months to a year. In this study, we propose a practical method for detecting OWTs from a single SAR image by integrating morphological features and spatial distribution analysis. The method combines adaptive target extraction, morphological filtering, and spatial statistical analysis to identify regularly spaced turbine clusters and derive their geographic coordinates. Case studies in the southern Shandong Peninsula, Jiangsu coast, and German Bight demonstrate that the proposed method achieves reliable detection performance across diverse marine environments. Rather than replacing high-accuracy centralized global datasets or multi-temporal deep learning systems, the proposed method serves as a lightweight alternative strategy, offering a practical solution for timely and on-demand OWT inventory updating, offshore engineering site assessment, and maritime navigation safety management.