Neural networks and paper seismic interpretation

In seismic interpretation, picking of horizons in 2D and 3D datasets is one of the most important tasks after processing. In some cases this is made easier by the use of automated tracking algorithms but these are not always successful where horizons are weak and discontinuous. The use of Artificial Neural Networks for recognition and tracking of such horizons is shown to be successful and may also be combined with conventional techniques to provide the interpreter with a robust automatic tracking methods.