一种基于压缩感知的稀疏孔径SAR成像方法

 高分辨大场景合成孔径雷达(SAR)成像给数据存储和传输系统带来沉重负担.本文针对条带式SAR成像,提出一种基于压缩感知技术的稀疏孔径SAR成像方法.该方法沿方位向以部分子孔径采样的方式获取降采样的原始数据,然后在距离向采用传统匹配滤波方法实现脉冲压缩处理,在方位向则利用小波基作为场景散射系数的稀疏基,并通过求解最小l₁范数优化问题重构方位向散射系数.该方法在存在多普勒参数误差情况下,能够有效实现多普勒参数估计,具有良好稳健性.仿真和实测数据成像结果表明所提算法在方位向严重降采样条件下仍能够实现无模糊的SAR成像,具有较强的有效性与实用性.

An Imaging Method Based on Compressive Sensing for Sparse Aperture of SAR

High resolution and wide swath synthetic aperture radar (SAR) imaging increases the load of data transmission and storage severely.To mitigate this problem,a novel compressive sensing-based imaging method for sparse aperture of SAR is proposed.In the proposed method,firstly,partial sub-apertures data is sampled in the azimuth direction to reduce the raw SAR data.Secondly,the conventional matchedfilter is used to perform pulse compression in the range direction.Finally,the wavelet basis is used as a sparse basis to reconstruct the scattering coefficients by solving an l₁ minimization optimization.The proposed method can precisely estimate the Doppler parameters in the presence of the Doppler parameters errors.Even if very limited samples can be obtained in the azimuth direction,the proposed algorithm can produce the unambiguous SAR image.Simulated and real SAR data experiments demonstrate that the effectiveness and stability of the proposed algorithm.