Minimal Similarity Accumulation Attribute for Fault Enhancement

In this work we present a new method for seismic fault enhancement on volumetric grids called Minimal Similarity Accumulation (MSA). The seismic reflection data is transformed into a multi-dimensional amplitude space and a clustering procedure is applied in order to minimize the noise interference. Thereafter, we use an algorithm to compute the MSA for each voxel in the volume. This is done by applying an autoadaptable function that uses the voxel neighboring information. The MSA measurement globally highlights the fault regions presented in the original volume. To validate the proposed method we use the volume of the Netherlands offshore F3 block downloaded from the Open Seismic Repository. In order to assess the proposed method and illustrate it advantages, a set of vertical 2D slices of the seismic data are presented providing a comparison between our results and images manually interpreted by a geologist. Finally, we conclude that the proposed method is sufficiently accurate.