Detection scene analysis for high frequency radar

Suffering from multiform time-variant and fluctuant clutter in High Frequency Radar (HFR), uniform target detection, tracking and clutter mitigation methods will result in false alarm or missing alarm. To select optimized detection and clutter mitigation method in the region of interesting, a method maximizing the separability of the resultant classes in amplitude of range-Doppler map is used for segmentation in multi-level detection background. Statistical and qualitative analysis is operated on the result of clutter extraction. Real data indicated that the proposed method can extract clutter region and adapt for various echo spectral maps for HFR as well, which can offer more accurate prior knowledge for target detection and clutter mitigation.