Neural network for seismic horizon picking

A Hopfield neural network can solve optimization problems. We use a Hopfield net for seismic horizon picking. The peak position of each seismic wavelet corresponds to one neuron. We transform the constraints for detecting local horizon patterns and the constraints for extracting one horizon each time into the system energy function. From the theory of Hopfield nets, changing the values of neurons can decrease the energy. The system will be stable until the values of neurons are not changed. One horizon is extracted by using the algorithm each time. We remove the extracted horizon from the original seismic data and extract the next horizon until the last horizon is extracted. From experimental results in a bright spot, the picked horizons can match the visual inspection