Texture attribute analysis based on strong background interference suppression

Abstract

Seismic texture attributes are closely related to seismic facies and reservoir characteristics. However, when a strong reflection interface overlying or underlying one target layer exists, their seismic response will mask the seismic response of the target layer. In this case, it is difficult to use texture attributes to identify geologic structures and reservoirs in the target layer. We have adopted a novel method to analyze texture attributes based on suppressing strong background reflection interference. First, we use the difference between the seismic response of the lateral heterogeneous reservoir and the underlying or overlying strong reflection. Second, we use morphological component analysis to separate the poststack data set into two parts: the seismic response of the target lateral heterogeneous reservoir (e.g., a channel sand body) with a small spatial distribution and the underlying or overlying strong reflection interference (e.g., the seismic response of the stable sedimentary stratum) with a wide spatial distribution. Then, we apply the texture attribute analyzing algorithm based on the voxel cooccurrence matrix to the seismic response of the target lateral heterogeneous reservoir for identifying covered structures and characterizing the reservoir. Finally, we apply the adopted method to a 2D synthetic data set and a 3D real field data set to evaluate the effectiveness of our method. The results of texture attribute analysis indicate that our method provides more detailed structural characterization and useful information.