基于方向金字塔分解与均值滤波联合的叠前地震数据随机噪声压制

2012年 51卷 第No. 4-期
阅读:69
查看详情
Random noise attenuation on prestack seismic data by joint application of steerable pyramid decomposition
and mean filter
1.成都理工大学地球物理学院,四川成都610059;2.成都理工大学油气藏地质及开发工程国家重
点实验室,四川成都610059;3.成都晶石石油科技有限公司,四川成都610041
Wang Jiang, College of Geophysics, University of Technology, Chengdu 610500, China
在方向金字塔多尺度、多方向分解原理的基础上,提出了基于局部拉普拉斯概率密度模型和
最大后验概率估计的随机噪声压制方法,该方法能够压制大部分随机噪声,但信号局部存在畸变
和残余噪声。均值滤波具有图像平滑的作用,可以消除信号畸变,压制残余噪声,因此提出了方
向金字塔分解与均值滤波联合的随机噪声压制方法。数值模拟实验和实际应用效果表明,该方法
能有效压制随机噪声,提高地震资料的信噪比。
Based on the principles of steerable pyramid decomposition in multi-scale & multi- direction, the random
noise attenuation method on the basis of local Laplace probability density model and maximum posteriori
probability estimation was proposed. The method can suppress most random noise, but introduces the
distortion and residual noise in local signals. Since the mean filter can smooth maps, it can eliminate the
signal distortion and suppress the residual noise. So we jointly applied the two methods to attenuate
random noise. Numerical simulation and practical application results show that the new method can
effectively suppress random noise and improve the signal-to-noise ratio of seismic data.
方向金字塔; 均值滤波; 拉普拉斯概率密度函数; 叠前随机噪声衰减;
steerable pyramid; mean filter; Laplace probability density function; prestack random noise
attenuation
;
10.3969/j.issn.1000-1441.2012.04.004