小波-维纳组合滤波算法及其在InSAR干涉图去噪中的应用

Wavelet-wiener combined filter and its application on InSAR interferogram

为了提高InSAR干涉图的滤波质量,在分析小波变换和维纳滤波各自优势的基础上,提出并构造了一种小波-维纳组合滤波器,实现了相应的滤波算法并开发了一套计算程序。为验证该算法的功效,选取美国Phoenix局部地区作为实验区域,使用ERS-1/2C波段干涉图作为滤波原数据,以视觉效果、相位导数标准偏差、奇异点个数以及数字高程模型精度作为评价指标,并与其他两种典型滤波算法即小波软阈值法和Goldstein法进行了比较,证实了小波-维纳组合滤波算法在干涉图去噪、保护边缘信息和精度等方面具有明显的优势。

In order to raise signal-to-noise ratio （SNR） of InSAR interferograms, this paper proposes a Wavelet-Wiener combined （WWC） filter in view of the respective merits of Wavelet transform and Wiener filter. The WWC algorithm and its computer program are developed to raise SNR of interferograms. To validate the proposed filter, the localized area around Phoenix in Arizona of USA is selected as the testing site and the ERS-1/2 C-band interferogram is utilized as the source data for filtering. Several indicators, including visualization effect, standard deviation of phase derivatives, number of residues and accuracy in DEM derived interferometrically, are taken into account to assess the effectiveness of this filter. The tested results show that WWC filter has some prominent advantages in terms of denoising, edge protection and improving DEM accuracy, if compared with two typical approaches presented previously, i. e., Wavelet soft-threshold filter and Goldstein filter.