基于正交频谱重构的微波三维成像方法

传统微波三维成像方法的回波数据量非常大, 基于目标场景稀疏的压缩感知成像方法虽然可以降低采样率和数据量, 但字典矩阵内存占用巨大, 且对连续分布目标成像效果不佳。针对上述问题, 本文分析了目标回波在三维频域的数据分布特征, 根据构建的频谱正交投影重构模型实现了目标三维频谱的重构。为了进一步优化重构模型, 以最小图像熵作为判别准则对重构的正则化参数进行了最优估计, 并得到最优的频谱重构结果。本文所提方法具有较好的成像效果, 较高的运算速度和较小的内存占用。计算机仿真实验和微波暗室实验验证了本文所提方法的有效性。

Microwave 3D imaging method based on orthogonal spectrum reconstruction

The size of echo signal data of the tranditional microwave three-dimensional imaging system is extremely high. Compressed sensing imaging methods based on sparse representation of target scene can reduce the sampling rate and data size, but the dictionary matrix takes a lot of memory and has a poor imaging quality for continuously distributed target. For these problems, this paper constructs an orthogonal projection wavefront reconstruction model to reconstruct the target three-dimensional spectrum, which based on the analysis of the echo data's characteristics in three-dimensional frequency domain. To further optimize the reconstruction model, the minimum image entropy is used as the criterion to optimum estimate the regularization parameters of reconstruction. The proposed method has the advantages of higher computing speed, less memory space, and better imaging quality for the continuously distributed target. The computer simulation experiment and microwave anechoic chamber measurement experiment verify the effectiveness of the proposed method.