【摘要】 基于双程波的逆时偏移会产生低频成像噪声,在成像后运用拉普拉斯滤波法可以取得较好的压噪效果,但是,该方法严重依赖于角度参数,使得滤波后的成像剖面上常常存在同相轴不光滑和噪声压制不完全的情况。基于叠后地震资料同相轴结构的特点,为进一步提高逆时偏移叠加数据的信噪比,引入了非局部平均滤波法,针对拉普拉斯滤波后的成像数据进行进一步的压噪处理。非局部平均滤波法将输入的地震数据分解为不含噪声的地震数据和噪声数据两部分,利用不同成像点与其它成像点间的相似系数,实现滤波处理。二维复杂模型VSP正演模拟数据逆时偏移结果的试算表明,应用非局部平均滤波后的剖面信噪比得到进一步提高,同相轴的连续性明显增强。实际VSP资料逆时偏移低频噪声压制试处理也表明非局部平均滤波方法具有计算精度高、算法稳定性好和易于实现的特点。更多还原

【Abstract】 Low-frequency noise is the inherent property of reverse-time migration(RTM)based on two-way wave equation.Filtering after imaging is widely used to suppress this kind of noise,especially Laplacian filtering method.But the result of Laplacian method depends on angle parameters,the case of rough events and residual noises always exist.In order to improve signal to noise ratio of imaging data of reverse-time migration,Non-local Means(NLM)filtering approach is introduced in this paper,which is used to suppress noise based on the characteristics of event structure of post-stack seismic data,and is applied to the seismic data after Laplacian filtering.By using NLM method,the input seismic data can be divided into noise and data,and the similarity coefficient of different imaging points is applied to realize filtering.For 2Dcomplicated theoretical model,we carried out forward modeling and RTM on VSP data.The results demonstrate that the combination of Laplacian and NLM filtering is better than the result of Laplacian filtering.Moreover,the application of actual VSP seismic data also suggests NLM filtering has advantages in computational accuracy,stability and practicability.更多还原