【摘要】 传统的F-X域预测滤波方法在理论上需要假设地震同相轴呈线性,当假设条件不满足时,在分离有效信号和随机噪声时会产生信号畸变。本文提出一种基于F-X域预测滤波和全变分滤波的串行压制噪声方法,对F-X域预测滤波的残差信号进行全变分滤波,并将全变分滤波结果与F-X域预测滤波结果相加作为有效信号的估计。对合成地震数据和实际地震数据的处理结果均表明,相对于F-X域预测滤波法,本文方法在压制随机噪声的同时能更好地保护有效信号,即显著提高信号的保真度。更多还原

【Abstract】 Random noise attenuation techniques are very important in seismic signal processing.They are used to improve seismic data SNR by extracting signals.The conventional F-X prediction filter assumes that the trajectories of the seismic events are linear.It will distort seismic signals if the linear events assumption is not satisfied.In this paper,we propose a cascaded filter,which is based on F-X prediction and total variation(TV) filter,for random noise attenuation.In the proposed method,the error signal,which is the difference between the original data and the filtering result obtained by F-X prediction method,is filtered further by the TV filter.Then the results obtained by F-X prediction and TV methods are combined together as the final output.The results of the synthetic and real data experiments show that the proposed method achieves better performance on the signal preservation than the F-X prediction method does.更多还原