摘要
常规的随机噪声压制方法面临着噪声频带与有效信号频带重叠,在压制噪声的同时对有效信号造成损害的局限性,基于小波变换和奇异值分解的思路,提出了一种小波变换与奇异值分解相结合的去噪方法,以单道信号作为处理单元,通过小波变换得到小波系数矩阵,并对此矩阵进行奇异值分解,进而求得能够反应信号与噪声变化的奇异熵,根据奇异熵确定阀值,进行SVD重构小波系数矩阵,最后小波逆变换重构信号,达到去除随机噪声的目的。此种方法对满足高斯白噪和不满足高斯白噪条件的随机干扰,均有去除效果。经理论信号与相关实际资料的处理证明,这种小波变换与奇异值分解相结合的去噪方法有效而实用。
The conventional random noise suppression method faces the overlapping of the noise frequency band and the effective signal band,together with the limitation of the damage to the effective signal while suppressing the noise.Because of the limitations of conventional random noise suppression methods,using the idea of wavelet transform and SVD,a denoising method based on wavelet transform and SVD is proposed in this paper:the single trace as the processing unit,get a matrix formed by various scales by wavelet transform,then SVD is used to obtain the singular entropy of the signal and noise,the wavelet coefficients are selected according to the singular entropy,and reconstruct the signal to remove the random noise from seismic data.This method can eliminate the random interference which satisfies the Gauss white noise and does not satisfy the conditions of the white noise of the Gauss.Through the theoretical signal and the actual data processing,it is proved that the denoising method combining wavelet transform and SVD is effective and practical.
作者
宋林
SONG Lin(Sinopec Geophysical Research Institute, Nanjing 211103, China)
出处
《物探化探计算技术》
CAS
CSCD
2018年第6期735-740,共6页
Computing Techniques For Geophysical and Geochemical Exploration
