基于稀疏约束最优化的ISAR相位自聚焦成像算法

Novel Autofocusing Algorithm for ISAR Imaging Based on Sparse Constraint

本文提出了一种基于稀疏约束的ISAR方位自聚焦算法,能够应用于稀疏孔径ISAR成像中.该算法利用ISAR图像的稀疏特征建立最小1范数成像模型,并将相位误差作为模型误差.然后通过数值迭代的方式进行自适应相位误差估计,最终获得聚焦良好的ISAR图像.同时,成像代价函数的建立基于矩阵模型,有利于采用方位FFT和矩阵的Hardmard乘积操作进行快速求解.由于利用稀疏约束,该方法在低信噪比的条件下仍然能够取得良好的聚焦结果.基于仿真数据和实测数据的结果验证了本文算法的有效性.

In this paper, a novel autofocusing algorithm of inversed synthetic apemn＇e radar （ISAR） imaging based on sparse conslraint is proposed, which can be applied in sparse aperture ISAR imaging. In the scheme, taking the phase errors as model er- rors, the proposed approach exploits the sparsity prior of ISAR image to construct the minimum l-norm image formation. Then nu- merical method is adopted to realize adaptive phase error estimation while well-focused ISAR image can finally be obtained.Mean- while, the objective function of ISAR imaging is established based on matrix model, which can be conveniently solved using fast Fourier transform （FFF） and matrix Hardmard multiplication. Due to the utilization of sparsity restriction, the proposed approach can still be capable of performing well even in the case of low signal-to-noise ratio （SNR）. The experimental results using both simulat- ed data and measured data confirm the validation of the proposal.