基于多普勒谱特征分析的舰船长度估计

A New Method for Ship Length Estimation Based on Doppler Spectrum Analysis

当前,各国越来越重视发展和保护本国的海洋权益,利用宽带雷达获取舰船目标的高分辨距离像并进行舰船类型识别逐步成为研究热点。船长是舰船的重要特征信息,对于判定舰船的类型和威胁等级具有重要意义。由于海杂波的时变和复杂性,利用信噪比门限方法确定距离像中舰船目标的边界并估计船长是不稳健的。从舰船类大刚体目标与海表面的运动特征差异出发,详细阐述了基于高分辨距离像多普勒谱分析的舰船长度估计方法,对不含目标距离单元和含目标距离单元多普勒谱的特征差异进行分析。提出分别利用多普勒谱的归一化波形熵特征和距离滑窗内距离单元的多普勒中心频率的均方差特征判定距离单元中目标的有无,实验结果证明了所提方法的有效性。

The vessel recognition through high range resolution profiles obtained from wideband radar gradually becomes a hot research topic. Ship length as an important feature is significant in determining vessel types and corresponding threat levels. Time variation and complexity of sea cluster make it difficult to get an accurate estimate of ship length with those algorithms which determine ship boundaries based on the specific SNR threshold. In this paper new methods for ship length estimation based on the Doppler spectrum analysis are presented in consideration of the difference in motion characteristics between large rigid body targets and sea cluster. The normalized waveform entropy features of Doppler spectrum and the mean square error of Doppler center frequency in range sliding windows are utilized respectively to decide whether there exist scat- tering points from a vessel in a specific range bin of HRRP. The experiment results demonstrate the validity of the proposed methods.