一种基于小波相位分析的InSAR干涉图滤波算法  

An Algorithm of Filtering InSAR Interferogram Based on Wavelet Phase Analysis

作　　者：蔡国林[1] 刘国祥[1] 李永树[1] 

机构地区：[1]西南交通大学土木工程学院测量工程系,四川成都610031

出　　处：《测绘学报》2008年第3期293-300,315,共9页Acta Geodaetica et Cartographica Sinica

基　　金：国家自然科学基金项目(40774004;40374003)

摘　　要：为了提高InSAR干涉图的滤波质量,本文在考虑干涉信号在同一尺度内具有相关性的基础上,提出并实现了一种基于小波相位分析的干涉图滤波算法。为检验算法的有效性,选取美国Phoenix地区高信噪比和伊朗Bam地区低信噪比两个真实C波段干涉图进行了滤波实验,通过与中值滤波算法、小波-中值滤波算法比较,证实了小波相位滤波算法是有效的且具有一定的优势。研究表明,不论是高信噪比干涉图还是低信噪比干涉图,小波相位滤波算法都能有效抑制其噪声。In order to enhance the quality in filtering InSAR interferograms, this paper proposes a wavelet-phase filter （WPF） which is based on wavelet phase analysis and takes into account the correlation of interferometric signal on the same scale. The experiments are then carried out to validate the algorithm by filtering two C-band interferograms, i. e. , high signal to-noise ratio （SNR） interferogram over Phoenix, USA and low-SNR interferogram over Barn, Iran. In comparison with the two existing filters, i.e. , median filter and wavelet-median filter, WPF proves to be more effective and advantageous to some extent. The testing results show that WPF can successfully suppress noises not only to high-SNR interferograms, but also to Iow-SNR interferograms.

关 键 词：合成孔径雷达干涉测量 干涉图 小波相位滤波 相位导数标准偏差 

分 类 号：TP75[自动化与计算机技术—检测技术与自动化装置]