基于加权子空间拟合的干涉相位估计方法

Image Autocoregistration and Interferogram Estimation Using Weighted Subspace Fitting

提出复合导向矢量的概念,并通过对测量的信号子空间与复合导向矢量张成的子空间进行拟和,推导出一种基于加权子空间拟合思想的干涉相位估计方法,该方法具有自适应图像配准和降低相位噪声功能,因而可以在图像配准精度很差(可以允许达到一个分辨单元)的条件下准确地估计相应像素间的干涉相位。仿真和实验数据表明估计结果优于基于子空间投影方法的估计结果。

The concept of compound steering vector was proposed, and interferogram estimate method based on the weight subspace fitting was proposed through the subspace fitting of measured signal subspace and the subspace spanned by compound steering vectors. The proposed method can autoregister the remote sense image and estimate the interferometric phases, and the phase noise is reduced simultaneously, in the presence of large coregistration errors, even up to one pixel. The processing results from simulating data and real data of InSAR and InSAS show that the proposed method is better than the subspace projecting method.