Seismic Random Noise Attenuation Using Optimal Empirical Wavelet Transform With a New Wavelet Thresholding Technique

The most vital challenge in seismic signal processing is the attenuation of random noise in seismic data. Many attenuation methods are formulated to mitigate the random noise but fail to retain high accuracy. In this article, a hybrid methodology based on optimal empirical wavelet transform (OEWT), energy kurtosis mean filtering (EKMF), and new wavelet thresholding (NWT) technique, namely, OEWT-EKMF-NWT, is proposed to suppress the random noise in the seismic signals. In the proposed OEWT-EKMF-NWT method, the OEWT decomposes the seismic signal into several intrinsic mode functions (IMFs) using the segmentation method. The mountain gazelle optimizer (MGO) algorithm with correlation waveform index (CWI) as an objective function is used here to choose the segmentation method optimally. Then, the EKMF is applied to select relevant optimal IMFs, which partially suppress the random noise. Furthermore, an NWT is proposed based on the discrete wavelet transform (DWT) to the relevant optimal IMFs. Here, an adaptive threshold calculation is designed based on the wavelet coefficients at different decomposition levels, which suppresses the random noise effectively. Finally, experiments are conducted on synthetic and real seismic signals to assess the efficacy of the proposed OEWT-EKMF-NWT method. The results reveal that the proposed OEWT-EKMF-NWT method attains better performance than the existing denoising methods.