NOISE ATTENUATION USING ADAPTIVE WAVELET THRESHOLD BASED ON CEEMD INF-X DOMAIN

Noise attenuation plays an important role in seismic signal processing. Complementary Ensemble Empirical Mode Decomposition (CEEMD) is a classic algorithm for signal decomposition and is usually used for denoising. This algorithm is used to attenuate random noise by removing some high-frequency intrinsic mode functions (IMFs), apparently resulting in insufficient noise attenuation and loss of effective signal. Wavelet threshold denoising can be used to attenuate the useless part and enhance the useful part of the signal by selecting the appropriate threshold. Wavelet threshold denoising is often combined with CEEMD in time domain to achieve relatively good effects, but some of signal between seismic traces are fragmented. This paper proposes improved adaptive wavelet threshold denoising based on CEEMD in f-x domain. The new threshold function we proposed is constructed on the basis of the traditional soft and hard threshold functions, which overcomes the constant deviation and avoids the phase step phenomenon. The processing results for simulated and field data show that the proposed method has better attenuation effect on random noise than traditional methods.

Automated summary

Noise attenuation is important in seismic signal processing. The classic algorithm, Complementary Ensemble Empirical Mode Decomposition (CEEMD), is commonly used for denoising, but it often results in insufficient noise attenuation and loss of effective signal. Wavelet threshold denoising can enhance the useful part of the signal by selecting the appropriate threshold, but it may cause signal fragmentation. This paper proposes an improved adaptive wavelet threshold denoising method based on CEEMD in f-x domain, which overcomes the limitations of traditional methods and shows better attenuation effect on random noise.