高分辨雷达一维距离像稀疏表示技术

Sparse representation technique for high range resolution profiles

高分辨雷达回波信号处理和特征提取技术受到广泛的关注。针对雷达目标几何绕射参数化模型,采用稀疏成份分析方法研究了高分辨距离像的超分辨重构和特征提取问题,分析了正则化函数的指标参数对信号重构性能和计算复杂性的影响,在此基础上提出了一种具有稀疏表示能力的正则化函数,实现了目标散射中心的有效分离。数值结果也验证了该方法的合理性和有效性。

Signal processing and feature extraction of high-resolution radar echoes have received intensive attention in recent years. Based on the parametric diffraction model of radar target, the method of sparse component analysis is applied to reconstruct the radar range profiles with super-resolution and their extract their features. The influences of the index parameter of regular function on the performance of signal reconstruction and computational complexity are analyzed. On this basis, a regular function with the ability of sparse representation is presented, by which the scattering centers of radar target are effectively separated while the amount of data storage is reduced. Accordingly its rationality and validity are proved by the numerical results.