A Machine Learning Based First-Order Sea Clutter Region Extraction Method for HFSWR

The first-order sea-clutter region extraction method is one of the key technologies for high frequency surface wave radar (HFSWR) for sea state remote sensing. However, the first-order sea clutter in the Range-Doppler (RD) map is hard to extract, especially in the presence of targets and splitting Bragg peaks. This paper presents a method to extract sea clutter based on machine learning with discriminative features from different aspects. The performance of the proposed method is verified and results demonstrate the effectiveness of the proposed method.