论文部分内容阅读
本文在对频率─空间域预测滤波压制随机噪声进行分析的基础上，实现了时间─空间域线性预测压制随机噪声。与频率─空间域方法相比，时间─空间域预测法可以自由地控制算子在时间方向的长度，尤其对中、低信噪比地震资料具有更强的压制随机噪声的能力。该方法可以很方便地扩展为全三维随机噪声衰减。文中给出了应用此方法的理论模型和实际资料的例子，表明它是一种很有效的去噪手段。
Based on the analysis of the random noise suppression in the frequency-space domain prediction filtering, the linear prediction of time-space domain is implemented to suppress the random noise. Compared with the frequency-space domain method, the time-space domain prediction method can freely control the length of the operator in the time direction, especially for moderate and low SNR seismic data with stronger suppression of random noise. The method can be easily extended to full three-dimensional random noise attenuation. The paper gives an example of the theoretical model and actual data applying this method, which shows that it is a very effective denoising method.