基于广义导向矢量的干涉合成孔径声纳干涉相位估计方法

提出了一种基于矩阵拟和的方法来估计干涉合成孔径声纳（InSAS）的干涉相位，该方法首先构造广义导向矢量，并利用相邻像素对的信息估计协方差矩阵，然后构造一个协方差矩阵全局最优的代价函数，使得这个代价函数最小时的相位即为精确的干涉相位。该方法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。通过仿真和InSAS试验数据验证了该方法的有效性和稳健性。

An estimation method of interferometric phase based on generalized steering vector

A method based on matrix fitting to estimate the interferometric phase of synthetic aperture sonar interferometry（InSAS） is proposed. In this method, generalized steering vector is formulated at first. And with the use of coherence information of neighboring pixel pairs, covariance matrix of generalized steering vector is estimated. Then a cost function of covariance matrix global optimum is given, which can estimate the exact interferomentric phase when the phase make the cost function get its minimum. The method can provide accurate estimate of the terrain interferometric phase （interferogram） even if the coregistration error reaches one pixel. The simulation and InSAS results show the validity and robust of this method.