基于矩阵填充的二维稀疏高分辨ISAR成像方法

Two dimensional sparse high-resolution ISAR imaging based on matrix completion theory

在SF ISAR成像过程中,由于外界环境以及雷达多工作模式的影响,回波矩阵中数据的缺失常会呈现随机性,从而导致传统CS方法无法直接进行处理。针对上述问题,提出一种基于矩阵填充理论的二维稀疏高分辨ISAR成像方法。首先,利用观测数据矩阵的低秩性质,将缺失数据的恢复问题转化为核范数最小化优化模型;然后,利用相应的矩阵填充优化算法进行求解;最后,在恢复的全数据基础上,直接利用二维FFT得出最终的高分辨率图像。与其它方法相比,该方法在低采样率和低信噪比条件下可以有效地抑制虚假重构,获得高分辨率图像,具有简单、易实现的优势。实测数据实验验证了该方法的有效性。

In the process of SF ISAR imaging,due to the influence of the external environment and radar multi-mode,the missing data in the echo matrix often presents randomness,which makes the traditional CS method unable to process directly.To solve these problems,a two-dimensional sparse high-resolution ISAR imaging method is proposed based on matrix filling theory.Firstly,by using the low rank property of the observation data matrix,the problem of missing data recovery is transformed into a kernel norm minimization optimization model;then,the corresponding matrix filling optimization algorithm is used to solve the problem;finally,on the basis of the recovered full data,the final high-resolution image is obtained by using two-dimensional FFT directly.Compared with other methods,this method can effectively suppress the false reconstruction under the condition of low sampling rate and low signal-to-noise ratio,and obtain high-resolution image,which has the advantage of simple and easy implementation.The experimental results of measured data verify the effectiveness of this method.