基于多小波自适应阈值的地震图像去噪方法
详细信息 本馆镜像全文    |  推荐本文 | | 获取馆网全文
摘要
结合地震图像噪声的特点,利用多尺度小波变换的优点,提出一种新的自适应小波阈值去噪算法。该算法根据小波变换各子层系数矩阵,确定各子层的自适应最优阈值,对于高频子带采用硬阈值化去噪,低频子带采用中值阈值化去噪,并去噪后重构小波系数。实验结果表明,该方法能够去除大部分高频随机噪声,并还原相干切片图像真实效果,提高了地震资料的信噪比。
Combined with the characteristics of seismic image noise,using multi-scale wavelet transform,a new adaptive wavelet threshold denoising algorithm is presented.This algorithm based on wavelet transform each layer,the coefficient matrix of the layer of adaptive optimal threshold,for high frequency sub-band adopts hard threshold denoising,low-frequency subband using threshold value denoising,after denoising and reconstruct wavelet coefficients.Experimental results show that this method can remove most random noise,and reductive real effect for coherent slices image,improve the signal-to-noise ratio.
引文
1西文,苏明军,朱文春,等.基于小波变换的地震相干算法及应用.石油物探,2003;41(3):334—338
    2Feris R S,Gemmell J,Toya-ma K,et al.,Hierarchical wavelet net-worksfor facial feature locallization.Fifth IEEE International Confer-ence on Automatic Face and Ges-ture Recognition,Washinton D.C,2005
    3陈善学,郑建宏.由小波变换模极大值实现信号重构.重庆邮电学院学报(自然科学版),2007;(2):12—15
    4谢荣生,李汉杰,孙枫,等.基于多小波噪声方差阈值的信号滤波方法.哈尔滨工程大学学报,2002:23(2):51—54
    5Donoho D L,Johnstone I M.Wavelet shrinkage:asymptopia.J R Stat Soc B,1995;57:301—369
    6林椹尠,宋国乡,薛文.图像的几种小波去噪方法的比较与改进.西安电子科技大学学报(自然科学版),2006;31(4):627—628

版权所有:© 2023 中国地质图书馆 中国地质调查局地学文献中心