高阶拉普拉斯算子逆时偏移低频噪声去除方法

叠前逆时偏移被认为是目前最为准确的成像方法，但是因为采用了双程波动方程和时间一致性成像准则，逆时偏移成像会产生很强的低频噪声。逆时偏移的成像噪声具有区别于有效波的低频特性、方向特性和大角度特性，这些特性是逆时偏移低频噪声去除方法的实现基础。拉普拉斯算子滤波去噪方法可以直接在偏移后的数据上进行，实现起来比较简单，增加的计算量小，在实际中应用较多;但是常规的拉普拉斯算子滤波去噪不彻底，而且改变了成像数据的振幅和相位信息。为了改进逆时偏移低频噪声去除效果，分析了拉普拉斯算子去噪的基本原理，研究了改进的高阶拉普拉斯算子去噪方法，修正了常规方法中存在的上述问题，既保证了去噪的效果，又保持了成像数据振幅和相位信息的相对不变。模型测试结果证明了高阶拉普拉斯算子逆时偏移低频噪声去除方法的可行性和有效性。

Prestack reverse-time migration is regarded as the most precise imaging method nowadays.However,due to adopt two-way wave equation and the time consistency imaging condition,there is strong low-frequency noise aroused in reverse-time migration.The noise in reverse-time migration is characterized by low-frequency,directional and large angle,which form the basis for reverse time migration denoising.Laplacian operator filter denoising method can be applied directly on the migration data,and it is easy to use with little computation cost.However,the traditional Laplacian operator denoising method has limitations in removing noise and does not take amplitude and phase into account.In order to improve the denoising effect,the basic theory of Laplacian denoising is firstly analyzed,and then an improved higher-order Laplacian operator denoising method is proposed.The proposed method modifies the existing problems in traditional method,the denoising effect is improved and the amplitude and phase of the imaging data are kept well.The feasibility and validity of this method is proved by model test.