Full text loading...
-
Multiscale Pre-Stack Seismic Attribute Enhancement Using Radial Basis Function Network
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, 80th EAGE Conference and Exhibition 2018, Jun 2018, Volume 2018, p.1 - 5
Abstract
The seismic attribute analysis has been widely applied in recent years as one of the effective approachs of reservoir prediction. However, traditional seismic attribute with full-azimuth stack is difficult to meet the requirement of reservoir development. In the paper, we propose a supervised multiscale attribute enhancement algorithm for pre-stack seismic data, which uses radial basis function network and downsampling pyramid. The method mainly composes of three steps: partial stack, multi-scale decomposition and reconstruction using neural network. In the applications to the 3D seismic data from Western China, the proposed method performs better in the seismic attributes enhancement and noise-suppression than classical post-stack seismic attributes (especially geometric attributes, such as coherence and curvature). Experimental results show that our method is effective in automatically improving the accuracy for the recognition, location and interpretation of reservoirs characteristics.