Joint Ship Detection Based on Time-Frequency Domain and CFAR Methods with HF Radar

Compact high-frequency surface wave radar (HFSWR) plays a critical role in ship surveillance. Due to the wide antenna beam-width and low spatial gain, traditional constant false alarm rate (CFAR) detectors often induce detection probability. To solve this problem, joint algorithm based on time-frequency (TF) analysis CFAR method is proposed paper. After TF ridge extraction, performed test each sample of ridges, binary integration run determine whether entire ship. verify effectiveness algorithm, experimental data collected by Ocean State Monitoring Analyzing Radar, type SD (OSMAR-SD) were used, with records from an automatic identification system (AIS) used as ground truth data. The processing results showed that TF-CFAR outperformed detecting non-stationary weak signals those within first-order sea clutters, whereas identifying multiple similar frequencies. Notably, intersection matched sets alone was not immense, which takes up approximately 68% number 25% TF-CFAR; however, union much (>30%) greater than result either method. Therefore, can further increase probability greatly improve performance under complicated situations, such non-stationarity, signal-to-noise ratio (SNR), clutters.