Predictive painting of 3-D seismic volumes

Conclusions

I have introduced predictive painting, a numerical algorithm for automatic spreading of information in 3-D seismic volumes according to the local structure of seismic events. The structure is extracted using plane-wave destruction, which operates by predicting each trace in the volume from its inline and crossline neighbors. In the second step, prediction operators are used to spread information inside the volume. This procedure is automatic and computationally fast because it requires only a small fixed number of operations per each trace or data sample. Synthetic and field data tests demonstrate the effectiveness of predictive painting in accomplishing such tasks as automatic flattening and horizon picking.

Further research should concentrate on combining automatic tools with interactive interpretation to allow the information extracted from seismic data to be integrated with the geological insight of the interpreter.