一种容积信息滤波局部迭代相位解缠方法

A local iterative phase unwrapping method based on cubature information filter

本文针对低信噪比干涉图相位解缠难题,提出了一种容积信息滤波局部迭代相位解缠方法。先采用基于修正矩阵束模型的局部相位梯度估计算法获取待解缠像元局部窗口相位梯度信息,通过引入堆排序的质量图指引相位解缠路径,再利用容积信息滤波算法在待解缠像元局部窗口内逐像元递推进行一维状态估计,直至遍历局部窗口内所有相邻像元,即可获得待解缠像元的状态估计,在保持解缠效率的同时提高了相位解缠精度。模拟和实测数据结果证明了本文算法的有效性。

To solve the phase unwrapping problem of interferograms with low SNR,a local iterative phase unwrapping method based on cubature information filter is proposed.Firstly,a local phase gradient estimation algorithm based on amended matrix pencil model is used to obtain the phase gradient information in the local window centered at the pixel to be unwrapped.Then the phase unwrapped path is guided by introducing the quality map of the heap sort.And the cubature information filter is used to recursively perform onedimensional state estimation operation.The state estimate of the pixel to be unwrapped is obtained until all neighbours in the local window of the pixel to be unwrapped are traversed.The phase unwrapping accuracy is improved while maintaining acceptable unwrapping efficiency.The simulation and measured data results prove the effectiveness of the algorithm in this paper.