对地震图像进行保边滤波的非线性各向异性扩散算法  

An edge-preserving smoothing algorithm of seismic image using nonlinear anisotropic diffusion equation

作　　者：王绪松[1] 杨长春[1] 

机构地区：[1]中国科学院地质与地球物理研究所,北京100029

出　　处：《地球物理学进展》2006年第2期452-457,共6页Progress in Geophysics

基　　金：中科院重大创新项目(KZCX1-SW-18)资助

摘　　要：地震资料的噪音消除是地震资料处理的一个重要步骤,以往的地震资料滤波,都容易伤害地震资料的高频信号.原始地震图像属于线形纹理图像,地震同相轴表现为线条结构.基于偏微分方程的非线性各向异性扩散技术通过结构张量的特征向量控制扩散方向,利用扩散张量的特征值控制特征方向上的扩散量,对原始地震图像进行图像平滑处理,增强具有一致性的同向轴及其所反映的重要地质结构.The noise muffling for the seismic data is a significant step in the seismic data processing. During the past seismic data filtering, the high frequent information is vulnerable. The prime seismic image owns the line-like texture characters, and the seismic event appears to the line structure. Based on the partial differential equation, the nonlinear anisotropic diffusion technique could control the diffusion direction through the structure tensor＇s eigenvector, and could deal with the diffusion size through the structure tensor＇s eigenvalue, and thus smoothes the prime seismic image. This method can enhance the coherent events and some important geology structure reflected by those events.

关 键 词：地震图像 偏微分方程 各向异性扩散 一致增强性扩散 平滑滤波 边界保护 地层倾角 

分 类 号：P631[天文地球—地质矿产勘探]