针对聚束合成孔径雷达(SAR)成像中的海量数据问题,提出了一种基于压缩感知的聚束SAR回波数据压缩成像方法。首先对回波信号进行分析,利用波前匹配字典构造方法得到回波观测矩阵,其次通过构造二维离散余弦变换(2D-DCT)等效矩阵得到回波信号的稀疏变换矩阵,然后利用平滑L0算法重构得到观测场景的二维像。本文方法可以在采集远小于传统聚束SAR成像方法所需回波数据量的基础上实现准确成像,最后的仿真结果表明了方法的可行性和有效性。
The amount of echo data is huge in spotlight SAR imaging. To solve this problem, the imaging method with compressed Spotlight SAR echo data based on Compressed Sensing is proposed. First, the echo measurement matrix is designed by the waveform-matched rules. Second, the sparse transformed matrix of the echo data can be obtained by constructing the 2D-DCT equivalent matrix. Then the smoothing L0 algorithm is utilized to reconstruct the target scene. The amount of raw data collected by the proposed method is much less than that of the conventional spotlight SAR imaging method. Finally, the feasibility and effectiveness of the proposed method are shown by the simulation results.