Testing of Clustering Algorithms on Different 3D Seismic Models
J. Amtmann, C.G. Eichkitz, M.G. Schreilechner, D. Hofer, N. Gegenhuber and M.F. Jud
Event name: 79th EAGE Conference and Exhibition 2017
Session: Seismic Interpretation - Analytics and Machine Learning for Interpretation
Publication date: 12 June 2017
Info: Extended abstract, PDF ( 726.3Kb )
Price: € 20
In seismic interpretation, a big amount of data has to be handled to segment the data cube in zones and faults. In the conventional method, inlines, crosslines and seismic sections are interpreted to divide the geological zones on seismic reflectors and on seismic discontinuities. This segmentation is often guided by seismic attributes, wells and further geological information. The other approach of seismic interpretation is dividing seismic data by algorithms. One popular method to achieve an automatic segmentation is clustering of seismic attributes. There are several clustering algorithms available in all different kinds of scientific disciplines. Some are also already used in seismic interpretation. To get an overview of clustering algorithms and to understand the different kinds of algorithms a research study was done. Therefore, multiple algorithms were classified in a matrix and a workflow was created to test various algorithms on different synthetic 3D seismic data models and subsequently a test environment was founded to understand algorithms to use them for automatic or semiautomatic interpretation of seismic data.