A fast algorithm for coherency estimation in seismic data based on information divergence

Coherency measurements have proven to be an effective method for representing geological discontinuities such as faults and stratigraphic features in 3-D seismic data volumes. Unfortunately, application of the algorithm suffers from the limitation of computation cost. This paper describes a new fast method for efficient and robust coherency estimation in 3-D seismic data. The method is characterized by greatly increasing the computational efficiency based on information divergence, which uses a recursion method and defines a new criterion by information divergence to calculate the coherency, to avoid directly computing the eigenvalues of the covariance matrix. In contrast to other algorithms, this method possesses higher computational efficiency and better anti-noise ability than commonly used methods. We demonstrate the advantage of this method using real seismic data examples.