分布式星载干涉合成孔径雷达(InSAR)是获取高精度地面数字高程图的有效技术之一。叠掩和阴影是雷达成像的几何关系导致的一类SAR图像中普遍存在的现象。在InSAR信号处理中,叠掩和阴影区域的相位奇异性,导致相位无法正确滤波和解缠,大大增加了这些区域及邻近区域的高程反演及地面定位的误差。本文从叠掩和阴影区域的几何模型和信号模型出发,基于干涉信号的自相关矩阵特征值,采用基于信息论方法的信号源个数估计方法,将叠掩和阴影区域从正常InSAR相位区域中区分出来,保证了后续处理步骤的有效性和可靠性。理论分析和仿真结果证实了本文方法的有效性和实用性。
Distributed spaceborne single-baseline InSAR is an effective technique to get high quality Digital Elevation Model. Layover and Shadow are ubiquitous phenomenon in SAR images because of geometric relation of SAR imaging. In the signal processing of single-baseline InSAR,the phase singularity of Layover and Shadow leads to the phase difficult to filtering and unwrapping. This paper analyzed the geometric and signal model of the Layover and Shadow fields. Based on the interferometric signal autocorrelation matrix,the paper proposed the signal number estimation method based on information theoretic criteria,to distinguish Layover and Shadow from normal InSAR fields. The effectiveness and practicability of the method proposed in the paper are validated in the simulation experiments and theoretical analysis.