| Seismic exploration technology is one of the most widely used means in the field of oil and gas exploration.However,in recent years,with the continuous in-depth development of exploration,the geological situation faced by the exploration area has become more and more complex and changeable.The collected seismic data has transitioned from the past narrow azimuth and low coverage times to "two widths and one height",which contains more anisotropic information and noise interference mixed with sea volume.This requires geologists to do more accurate processing of the collected seismic data to reflect the most real geological structure and lithology information.On the basis of not damaging the effective seismic signal energy,various noise interferences are suppressed to the greatest extent,that is,the "amplitude preserving denoising" in seismic data processing.Curvelet Transform technology is widely used in the "amplitude preserving denoising" step of seismic data processing because of its excellent ability of multi-scale,multi-directional and characterizing the characteristics of local edge signals,and has achieved good results.Based on the theory of Curvelet Transform,this paper studies the application in suppressing seismic random noise and surface wave interference.Although the traditional soft,hard and linear Curvelet threshold denoising methods have good suppression effect on random noise,the suppression of each scale of the original noisy signal in Curvelet domain is holistic,and the selection of threshold is single,which will damage some effective signals.In view of this situation,this paper introduces the Curvelet threshold denoising method based on BEMD,which improves the effect of suppressing random noise.The effectiveness of this method is verified by comparing the wide azimuth denoising method with the actual seismic data.Aiming at the problem that the conventional Curvelet transform method divides the spectrum too single in the process of suppressing surface wave and removes the low-frequency effective signal overlapping with surface wave interference,a method based on Empirical Curvelet Transform is proposed in this paper.Empirical Curvelet Transform can adaptively identify the division scale and direction according to the distribution of effective reflected signal and surface wave interference in Curvelet domain,so as to divide the spectrum more accurately,which is conducive to the accurate suppression of noise coefficient in the later stage.Through the analysis and comparison between the forward model and the actual wide azimuth seismic data from the aspects of visual effect and signal amplitude preservation,it can be found that the processing effect of this method is more obvious than the traditional Curvelet Transform surface wave suppression method,and the purpose of "amplitude preservation and denoising" of seismic signal is realized. |
