GHM类正交多小波变换及其在地震资料去噪中的应用  

GHM-like orthogonal multiwavelet transform and its application in noise removal for seismic data

作　　者：陈香朋[1] 吕小伟[1] 曹思远[2] 

机构地区：[1]胜利油田有限公司物探研究院 [2]石油大学(北京)物探重点实验室

出　　处：《油气地球物理》2005年第3期15-19,共5页Petroleum Geophysics

摘　　要：多小波是对小波理论的一个新发展,它可以同时满足正交性、对称性、短支撑等良好的特性要求。本文介绍了多小波变换的基本理论、变换过程及预处理方法,提出了基于GHM类多小波分析的地震信号去噪方法。对理论数据和实际资料进行的软阈值去噪处理表明,多小波变换在有效压制随机噪声的同时,还能较好地保留原信号的特征信息,是一种行之有效的去噪方法。Multiwavelet is a new development to the wavelet theory and it simultaneously offers orthogonality, symmetry, and short support. The paper introduces multi-wavelet principles, transform process and pre-processing methods. Based on GHM-like multiwavelet transform, a new method for noise reduction of seismic data is presented in the paper. The experiments of noise reduction to analog and real data show that multiwavelet transform is useful for noise reduction and can preserve signal features at the same time.

关 键 词：GHM类正交多小波变换 地震资料 去噪处理 地震勘探 地震信号 对称性 短支撑性 二阶消失矩 正交性 

分 类 号：P631.44[天文地球—地质矿产勘探]