摘要: |
针对传统压缩感知(Compressive Sensing,CS)重构算法成像精度低及抗噪性能差等问题,提出了一种基于自适应加权极小极大凹罚函数和全变分的稀疏合成孔径雷达(Synthetic Aperture Imaging Radar,SAR)成像重建方法。首先,将加权思想同非凸函数簇中的极小极大凹罚函数结合,以进一步促进解的稀疏性;然后,与全变分判罚函数线性组合构成复合正则化器,以进一步提高抗噪性能;最后,采用交替方向乘子法求解该成像模型,并在求解过程中使用方位-距离解耦算子替换测量矩阵及其厄米特转置以减少存储空间。仿真与实测数据处理结果表明,所提方法相比于其他算法有更好的聚焦性能和重建精度。 |
关键词: 稀疏SAR成像 自适应加权 非凸正则化 全变分 极小极大凹罚函数 |
DOI:10.20079/j.issn.1001-893x.220418001 |
|
基金项目:国家自然科学基金资助项目(62071238);江苏省自然科学基金(BK20191399) |
|
A Sparse SAR Imaging Method Based on Adaptive Nonconvex Regularization and Total Variation |
LI Jiaqiang,HU Zhangyan,YAO Changhua,GUO Guixiang,CHEN Jinli |
(School of Electronic and Information Engineering,Nanjing University of Information Science and Technology,Nanjing 210044,China) |
Abstract: |
For the problems of low imaging accuracy and poor anti-noise performance of traditional compressive sensing reconstruction algorithms,the authors propose a sparse synthetic aperture radar(SAR) imaging reconstruction method based on adaptive weighted minimax concave penalty function and total variation.Firstly,the weighted idea is combined with the minimax concave penalty function in the non-convex function cluster to further promote the sparsity of the solution.Then,the composite regularizer is constructed by linear combination with the total variational penalty function to further improve the anti-noise performance.Finally,the alternating direction multiplier method is used to solve the imaging model.The measurement matrix and its Hermitic transpose are replaced by azimuth-range decoupling operator to reduce the storage space.Simulation and experimental data processing results show that the proposed method has better focusing performance and reconstruction accuracy than other algorithms. |
Key words: sparse SAR imaging adaptive weighting non-convex regularization total variation minimax concave penalty |