Target Detection Based on HSV Feature Fusion and Time-Frequency Analysis for HFSWR

High frequency surface wave radar (HFSWR) is capable of beyond-the-horizon surveillance of vessels in vast areas. However, the complex marine environment introduces various clutter and interference that are detrimental to the detection of HFSWR targets. In this paper, a target detection algorithm for sea clutter regions is proposed based on time-frequency analysis(TFA). The algorithm presents a multi-stage detection framework, initially identifying suspicious target positions in clutter through range-doppler (RD) spectrum analysis, followed by TFA of the echoes at these positions, feature fusion with manual features, and final classification using a deep neural network. Compared to classical target detection algorithms, this method exhibits superior detection performance in sea clutter regions.