【摘要】 在非线性各向异性扩散滤波基础上,采用Jacobi算法求解结构张量D中的特征分量的特征值u,并建立线状置信度C_line求解扩散滤波系数λ,将特征值u和扩散滤波系数λ代入扩散方程来求解该方程,实现了置信扩散滤波方法,该方法的理论不同于常规基于结构张量的扩散滤波方法。通过对加噪的理论模型进行试验,证明了该方法的有效性,并对比带通滤波方法进行频谱分析,同时利用SNR（信噪比）、MSE（均方误差）曲线分析了迭代次数对处理结果的影响。最后将该方法应用于实际叠前和叠后地震资料,较好地衰减了随机噪声、保留了地震有效信号,有效地提高了地震资料的信噪比。该方法相比常规基于结构张量的扩散滤波方法,对噪声有更好压制效果。更多还原

【Abstract】 The method on the basis of nonlinear anisotropic diffusion filtering, can use Jacobi algorithm to solve the eigenvalue of the eigen components in the structure tensor D. A linear confidence measure Cline is built to solve the diffusion filter coefficients λ. The eigenvalue and the diffusion filter coefficients λ are substituted into the diffusion equation to solve the equation, which can realize the confidence diffusion filtering method. The theory of this method is different from the conventional diffusion filtering method based on structure tensor. Processing a theory record of including random noise model, proves the validity of the method, and analyzes the frequency spectrum by comparing the band-pass filtering method. On the other side taking advantage of the curves of SNR(signal-to-noise ratio) and MSE(mean square error), the relationship between the number of iterations and the results are analyzed. Finally, this method which is applied to real pre-stack and post-stack seismic data, can attenuate the random noise, retain the effective seismic signal and effectively improve the SNR of seismic data. This method has better noise suppression effect than the conventional diffusion filtering method based on structure tensor.更多还原