基于误差迭代补偿的干涉合成孔雷达四向加权最小二乘相位解缠法

摘要

Phase unwrapping is one of the key steps of terrain elevation measurement in interferometric synthetic aperture radar (InSAR).Four directions least-square algorithm,comparing the two directions weighted least-square algorithm,takes full consideration of the diagonal direction of the phase information in order to acquire solution results closing to the true value,but there' s still some errors.In order to further improve unwrapping accuracy,a new four directions weighted least-square method using errors iterative compensation is proposed.The phase gradient correlation based quality map and second order difference of wrapped phase are utilized to construct the new weight,which can better correct the phase region of bigger slope changes and offset the smooth effect.This method acquires higher unwrapping accuracy by error iterative compensation.The experimental results show that the method has certain feasibility and superiority.%相位解缠是干涉合成孔径雷达(InSAR)高程地形测量中的关键步骤之一.四向最小二乘法相比传统的两向最小二乘法,充分考虑了对角方向的相位信息,其解缠结果更接近真实值;但还是存在一定的误差.为了进一步提高解缠精度,提出了一种基于误差迭代补偿的四向加权最小二乘相位解缠法.通过利用缠绕相位二阶差分和梯度相关质量图的双重特性,来构建新的权重,能更好地修正坡度变化大的相位区域,抵消平滑作用,再采用误差迭代补偿机制,进一步提高解缠精度.实验结果表明,该方法有一定的优越性和可行性.