基于短时傅立叶谱脊的相位图滤波算法及仿真研究  

Research on Simulation and Filtering Algorithm for InSAR Interferogram Based on Windowed Fourier Spectrum Ridge

作　　者：黄柏圣[1] 许家栋[1] 

机构地区：[1]南京电子技术研究所,南京210039

出　　处：《系统仿真学报》2009年第23期7612-7616,共5页Journal of System Simulation

基　　金：陕西省自然科学基础研究计划资助项目(2006F15)

摘　　要：提出了一种基于二维短时傅立叶谱脊的相位图滤波方法。该方法首先将干涉相位数据转变成指数,利用一种新的局部频率估计方法对其进行局部频率估计,求取行列方向的最小和最大频率估计值,确定频率取值范围,选择高斯窗口尺寸;然后,通过二维短时傅立叶变换对指数数据进行处理,提取频谱峰值处频率,求其对应的相位;最后,采用仿真数据和Etna火山干涉数据进行试验,将算法与均值和Lee滤波方法进行了比较。试验结果表明,该方法在有效抑制相干斑的同时,还能有效地保持相位的细节信息和条纹的边缘结构,而且清除了残余点,有利于提高干涉测量的精度。An interferometric phase noise reduction algorithm based on two-dimensional windowed fourier spectrum ridge（WFSR） was proposed.Firstly,the interferogram phases were transformed into an exponential field,the local frequency estimates were computed by the approximate maximum-likelihood（AML） frequency estimate method,by the local frequency estimates,the minimum and maximum frequency values in the row and column were obtained,the range of frequency value was determined,and the Gaussian window size was selected.Then the exponential data were processed by two-dimensional windowed fourier transform（WFT）,when the extraction frequencies maximized the amplitude spectrum,the spectrum ridge and phase values were computed.By using both simulated and Etna interferograms,the performance of the method was demonstrated and compared with the mean filter and the Lee filter.The result shows that the new method is powerful to interferogram speckle noise suppression,and residues are removed,as well as it can preserve fine phase details and fringes＇ structures.The method proposed can improve the accuracy of interferometric products.

关 键 词：短时傅立叶变换 仿真 干涉相位图 滤波 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]