基于奇异值分解的复杂转动目标成像

目标在三维空间复杂转动时，传统的逆合成孔径成像算法得不到聚焦的二维像。通过对复杂转动目标等效散射中心相位变化的分析，提出了一种基于奇异值分解的成像算法。该算法首先估计多个散射中心的相位历程，获得相位矩阵|然后对相位矩阵进行奇异值分解，并根据奇异值之间的关系对目标是在三维空间转动还是在二维平面内转动进行判断。根据判断结果分别进行处理：当目标在平面内转动时得到目标的二维像|当目标在三维空间转动时得到失真的三维散射中心模型。仿真及实测数据处理结果验证了所提算法的有效性。

Imaging for complex rotation target based on singular value decomposition

The traditional inverse synthetic aperture radar (ISAR) imaging algorithm cannot obtain a focused image when the target rotates complexly in three dimensional space. After analyzing the echo phase for the scattering centers of complex rotating targets, an imaging algorithm based on singular value decomposition (SVD) is proposed. Firstly, the phase histories of scattering centers are estimated and the phase matrix is gotten. After doing SVD for the phase matrix, the target’s rotation in the three dimensional (3D) space or in the two dimensional (2D) plane can be judged by the singular values. Afterward, these two situations can be dealt with respectively, that is to say, a focused 2D image can be obtained when the target rotates in the 2D plane and the distorted three dimensional target scattering center model can be obtained when the target rotates in the 3D space. Simulation and experimental results verify the effectiveness of the algorithm.