1. School of Opto-Electronic Information, University of Electronic Science and Technology of China, Chengdu, Sichuan 610054, China;
2. Sichuan Petroleum Geophysical Prospecting Company, Chuanqing Drilling Engineering Co, Ltd, CNPC, Chengdu, Sichuan 610213, China
Abstract:A novel method for seismic multi-attribute fusion based on Contourlet transform is presented in the paper. First the multi-resolution, multi-directionality, localized features and anisotropy of each seismic attribute can be effectively represented by Contourlet transformation. Then decomposed low-frequency sub-bands are fused with the sum modified Laplacian (SML), and high-frequency sub-bands are fused with weighted coefficients obtained by the correlation of each scale coefficients. This method can not only enhance edge features and details in seismic attributes, but also remove noise in the fusion. Experimental results show that the proposed method is superior to the classic wavelet transform and other conventional fusion methods. Real data applications indicate that this method can eliminate the redundancy of attributes and improve oil-gas reservoir prediction and fluid identification with full use of different attributes.