Application of fault likelihood attribute based on edge-preserved filter in complex fault identification of Xihu Sag

The N structure of the central anticlinal belt of Xihu Sag is rich in oil and gas resources, but it has undergone frequent geological activities with complex fault stages and cross-cutting relationship. In addition, the signal to noise ratio of seismic data in the target area is low, and the reflection of the target interval is chaotic, which further affects the fine description of faults. In this paper, nonlinear anisotropic edge-preserved filter technology was used to denoise and overcome the problem of traditional linear filter to fault edge, and considered the noise suppression and boundary information protection. The fault tendency and dip angle were scanned at each sample point, the similarity coefficient and likelihood attribute were calculated successively, and the maximum likelihood attribute value was selected as the final result. This method could characterize faults more accurately than the traditional coherence algorithm. The application of actual data shows that the combination of nonlinear anisotropic edge-preserved filter and fault likelihood attribute could better suppress random noise and protect fault boundary information, the fault characterization precision was greater, and could better characterize the complex fault stages in the N structure in more detail.