Premium
A method for classifying pre‐stack seismic data based on amplitude–frequency attributes and self‐organizing maps
Author(s) -
MolinoMineroRe Erik,
RubioAcosta Ernesto,
BenítezPérez Héctor,
BrandiPurata Juan Marcos,
PérezQuezadas Nora Isabel,
GarcíaNocetti Demetrio Fabián
Publication year - 2018
Publication title -
geophysical prospecting
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.735
H-Index - 79
eISSN - 1365-2478
pISSN - 0016-8025
DOI - 10.1111/1365-2478.12607
Subject(s) - stack (abstract data type) , geology , seismic to simulation , set (abstract data type) , data set , data mining , amplitude , seismology , metric (unit) , wavelet , computer science , cube (algebra) , synthetic seismogram , seismic inversion , pattern recognition (psychology) , artificial intelligence , mathematics , engineering , programming language , operations management , geometry , quantum mechanics , combinatorics , azimuth , physics
Analysis of pre‐stack seismic data is important for seismic interpretation and geological features classification. However, most classification analyses are based on post‐stack data, which ignores pre‐stack information, and it may be disadvantageous for complex geological description. In this work, we propose a method to address the classification of pre‐stack seismic data decomposed using the wavelet transform to spread the amplitude and frequency seismic attributes at the same time, which are then classified by a self‐organizing map. The resulting classes constitute an attribute constructed by the joint amplitude–frequency components of the transformed pre‐stack seismic gathers, which create a multi‐dimensional set defined through a given metric. Tests on a real seismic cube revealed that the method can identify patterns observed on the seismic images, which agree with our current knowledge of the seismic data. The method can be used as a complementary tool to identify features and structures in seismic signals.