Curvelet domain-based prestack seismic data denoise method
Zhang Heng-lei1, Zhang Yun-cui2, Song Shuang1, Liu Tian-you1
1. College of Geophysics and Space Information,China University of Geoscience,Wuhan City,Hubei Province, 430074,China;
2. GRI, Fifth Geophysical Prospecting Brigade of SINOPEC,Xiangtan City,Hunan Province, 411104,China
Abstract：The paper expounded idea of method for eliminating random noise of prestack seismic data on the basis of Curvelet domain. The processing of prestack seismic data is carried out by using non-linear threshold method in Curvelet domain through multi-scale decomposition and on the basis of curve changing feature having good approach. It is showed by numeric computation and processing of practical data in the paper that Curvelet domain-based denoise method can effectively suppress random noises on the seismic section: numeric computation showed minimum mean-square error and the highest S/N ratio, and processing of practical seismic sections showed good visual effect on the sections, which is characterized by clear and continuous events of reflections and provides as seismic data with higher S/N ratio for the following processing and interpretation such as prestack migration and AVO analysis etc.
