基于快速Myriad滤波方法的地震数据去噪

Myriad滤波是一种针对非高斯脉冲噪声而提出的非线性滤波方法。基本Myriad滤波方法的运算量受窗口长度的影响，窗口取得越长，运算量也越大;改进的快速Myriad滤波方法实现简单，执行效率高。在方法原理介绍的基础上，假设地震噪声服从α稳定分布，将快速Myriad滤波方法应用于地震数据的去噪处理试验，并分别与均值滤波方法和中值滤波方法进行了对比。理论模型数据测试和实际叠前地震资料应用的结果都表明，快速Myriad滤波方法不仅对脉冲噪声具有很强的抑制能力，而且能有效地保护数据中的有用信息，其性能远远优于同窗口长度的均值滤波和中值滤波方法，滤波后地震数据的信噪比明显增强。

Myriad is a kind of non-linear filter that is designed for non-Gaussian impulsive noise suppression.The computation amount of basic Myriad filter is influenced by the length of Myriad window.The longer the window is made,the greater the amount of calculation.Compared with the basic Myriad,the improved fast Myriad is easy to operate with high efficiency.In this paper,on the assumption that seismic noise follows α-stable distribution,the fast Myriad filter is applied to noise suppression experiment.A comparison among fast Myriad,mean filter and median filter is conducted on both theoretical model test and actual seismic data application,which show that the fast Myriad method has stronger ability to eliminate the impulse noises and preserve the desired information than mean filter and median filter,and its performance is much superior than the former two methods in the same window length.The S/N of the seismic data after Myriad filtering is highly improved.