机构地区: 湖南大学电气与信息工程学院
出 处: 《测绘学报》 2006年第2期128-132,共5页
摘 要: 提出一种基于小波变换的InSAR干涉图滤波算法,此算法先用小波变换对干涉图数据做多级分解,得到图像的多级近似部分系数和3个(水平、垂直、对角线)方向的细节部分系数,然后分别对每一级各个方向的细节部分系数检测其是否为对应方向的边缘,对边缘处的系数根据边缘的方向不同,用不同的方向模板平滑后,再中值滤波;对非边缘处的系数直接中值滤波。用真实的InSAR干涉图实验结果证明此方法具有较好的滤波效果。 A method for InSAR interferogram filtering based on wavelet transform is proposed. It first decomposes the InSAR interferogram with muhilayer using wavelet transform, and obtains multilevel approximate coefficient and detail coefficient with three orientations (horizontal, vertical and diagonal), then checks the detail coefficient of all the orientations at all levels, judges whether every coefficient is the edge of corresponding orientation or not, if yes, smoothes the coefficient using different mode corresponding to its orientation, then filters the coefficient using median filtering, otherwise filters the coefficient using median filtering directly. The experiment results using real InSAR interferogram show that this method is efficient.
领 域: [自动化与计算机技术] [自动化与计算机技术]