Key points are not available for this paper at this time.
Summary We present a clustering based methodology used to process 3D seismic data and map automatically a dense amount of horizons. It firstly transforms seismic reflection data from the amplitude space into multi-dimensional amplitude spaces. Then it uses different instances of a specific clustering procedure to create various distinct datasets of clusters. Each voxel is represented by a set of cluster labels, minimizing the interference of noise and uncertainties naturally present in the seismic data. The method finds the global structure of lateral amplitude reflections. It is capable of mapping a great portion of the seismic horizons present in the data, even when severely interrupted by seismic faults, salt domes and other discontinuities.
Figueiredo et al. (Tue,) studied this question.