曲波变换三维地震数据去噪技术

3D seismic data de-noising approach based on Curvelet transform

常规去噪方法主要用于二维地震资料的处理,但噪声也存在于三维数据空间,导致常规去噪方法效果较差。为此,选用多尺度、多方向的二维曲波变换进行三维地震资料去噪。首先抽取含噪三维地震数据的时间切片进行多尺度、多方向分解,在曲波域采用局部阈值法去噪,对曲波变换后的每一个尺度选取一个合适的阈值因子,然后通过阈值处理,得到各尺度下的有效波曲波系数,最后将提取的有效波曲波系数进行逆变换重构出地震信号,达到去噪的目的。理论模型与实际资料的处理结果表明,与其他去噪方法相比,该方法最大限度地保留了地震数据的有效信号,达到了更好的去噪效果,具有较好的应用前景。

Most conventional de-noising methods are mainly used for 2D seismic data. However noise develops also in 3D data space. So these conventional methods do not work well in 3D seismic data processing. Therefore we choose multi-scale and multi-directional 2D Curvelet transform to eliminate noise in 3D seismic data. During the de-noising process, the time slice extracted from 3D seismic data is decomposed by the multi-scale and multi-directional Curvelet transform, and the idea of local threshold de-noising is introduced. We choose a suitable threshold factor for each scale after Curvelet transform, and then we can get the Curvelet coefficient of signals. At last, we process the inverse transform to reconstruct seismic signals using the extracted Curvelet coefficient, and achieve the purpose of de-noising. Test results on theoretical model and real data show that the proposed approach can well preserve seismic signals and achieve better de-noising compared with other methods. © 2017, Editorial Department OIL GEOPHYSICAL PROSPECTING. All right reserved.