Weak Target Detection Based on the Average Bandwidth of the Partial Hilbert Spectrum
基于局部Hilbert谱平均带宽的微弱目标检测算法

In the paper, the partial Hilbert spectrum near the frequency of the target ridge and average bandwidth of the real sea clutter data is analyzed, and the influence of the target and polarization on the ridge and the average bandwidth of the partial Hilbert spectrum are discussed. As shown of the result, the appearance of the target will smooth its ridge and decrease its average bandwidth, on the base of the result, a novel method is proposed to detect the weak target in the sea clutter, where the average bandwidth is calculated by the partial Hilbert spectrum near the frequency of the target ridge to extract the difference of the target and the sea clutter, then it is used as the detection statistic to detect the weak target in the sea clutter. Compared with the target detection method directly using the box dimension and the Cell Average-Constant False Alarm Rate (CA-CFAR, 100 pulses), the proposed method improves the detection performance of the weak target in the sea clutter distinctly.