基于分布式压缩感知的全极化雷达超分辨成像  

Full polarization super-resolution radar imaging algorithm based on distributed compressive sensing

作　　者：吴敏[1] 张磊[1] 邢孟道[1] 段佳[1] 徐刚[1] 

机构地区：[1]西安电子科技大学雷达信号处理国家重点实验室,陕西西安710071

出　　处：《电波科学学报》2015年第1期29-36,共8页Chinese Journal of Radio Science

基　　金：国家自然科学基金优秀青年基金(No.61222108);国家自然科学基金(No.61101245)

摘　　要：基于分布式压缩感知理论,提出了一种全极化逆合成孔径雷达超分辨成像算法,联合各极化通道进行超分辨处理.首先,建立全极化信号模型及超分辨字典,利用各极化通道信号的联合稀疏性将全极化超分辨成像建模为最小L2,1范数的优化问题,运用一种快速算法求解该优化问题.由于利用联合稀疏约束,多极化通道联合成像相比于单通道成像能够获得更好的超分辨性能和噪声抑制能力,最终有效提高图像极化融合的效果.同时,采用快速傅里叶变换操作提升了算法的运算效率.基于backhoe的仿真数据实验验证了该算法的优越性.A novel super-resolution imaging algorithm for full polarized inverse synthetic aperture radar(ISAR)is addressed.Based on the distributed compressive sensing(DCS)theory ajoint processing of polarization and super-resolution is realized.The fully polarized signal model is established,based on which the super-resolution dictionary is formed.By exploiting the joint sparsity between polarimetric channel signals,the fully polarized super-resolution imaging problem can be mathematically converted into a L2,1norm optimization question.The optimization problem can be solved via fast optimization algorithm.Comparing with the single-polarization imaging,the jointly multi-polarization imaging performs better on super-resolution and noise suppression by utilizing joint sparsity.Besides,the efficiency of the proposed algorithm can be improved by fast Fourier transform(FFT).Simulated experiments of the backhoe data verify the effectiveness of the proposed algorithm.

关 键 词：逆合成孔径雷达 超分辨成像 分布式压缩感知 极化 

分 类 号：TN957.52[电子电信—信号与信息处理]