Abstract:Seismic data often undergo severe noise due to environmental factors, which seriously affect subsequent applications. Traditional hand-crafted denoisers such as filters and regularizations utilize interpretable domain knowledge to design generalizable denoising techniques, while their representation capacities may be inferior to deep learning denoisers, which can learn complex and representative denoising mappings from abundant training pairs. However, due to the scarcity of high-quality training pairs, deep learning denoisers may sustain some generalization issues over various scenarios. In this work, we propose a self-supervised method that combines the capacities of the deep denoiser and the generalization abilities of the hand-crafted regularization for seismic data noise attenuation. Specifically, we leverage the Self2Self (S2S) learning framework with a trace-wise masking strategy for seismic data denoising by solely using the observed noisy data. Parallelly, we suggest the weighted total variation (WTV) to further capture the horizontal local smooth structure of seismic data. Our method, dubbed as S2S-WTV, enjoys both high representation abilities brought from the self-supervised deep network and good generalization abilities of the hand-crafted WTV regularizer and the self-supervised nature. Therefore, our method can effectively remove noise and preserve the fine details of the seismic signal. To tackle the S2S-WTV optimization model, we introduce an alternating direction multiplier method (ADMM)-based algorithm. Extensive experiments on synthetic and field noisy seismic data demonstrate the effectiveness of our method as compared with the state-of-the-art traditional and deep learning-based seismic data denoising methods. Code is available at https://github.com/XuZitai/S2S-WTV.

Random Noise Attenuation Based on Support Vector Regression and Adaptive Wiener Filtering

Noise Reduction by Support Vector Regression with a Ricker Wavelet Kernel

Study of Least Squares Support Vector Regression Filtering Technology with A New 2d Ricker Wavelet Kernel

Random noise reduction using SVD in the frequency domain

Random Noise Attenuation Using a New Spectral Decomposition Method

An Adaptive Time-Frequency Denoising Method for Suppressing Source-Related Seismic Strong Noise

Self-Supervised Attenuation Method Based on Similarity Comparisons for 3D Seismic Random Noise

A Robust Random Noise Suppression Method for Seismic Data Using Sparse Low-Rank Estimation in the Time-Frequency Domain

An Effective Approach to Attenuate Random Noise Based on Compressive Sensing and Curvelet Transform

Weighted Multisteps Adaptive Autoregression for Seismic Image Denoising

An Adaptive High-Dimensional Progressive Denoising Method for Seismic Weak Signal Enhancement

Random Noise Attenuation by F-X Empirical-Mode Decomposition Predictive Filtering

Automatic noise attenuation based on clustering and empirical wavelet transform

Application of spectral decomposition using regularized non-stationary autoregression to random noise attenuation

Unsupervised Deep Learning for Random Noise Attenuation of Seismic Data

S2S-WTV: Seismic Data Noise Attenuation Using Weighted Total Variation Regularized Self-Supervised Learning

Support Vector Regression Based on Ricker Wavelet Kernel Function and Its Application to Seismic Prospecting Data Denoising

RANDOM NOISE REDUCTION USING A HYBRID METHOD BASED ON ENSEMBLE EMPIRICAL MODE DECOMPOSITION

Signal-Preserving Erratic Noise Attenuation Via Iterative Robust Sparsity-Promoting Filter

Structure-Oriented Singular Value Decomposition for Random Noise Attenuation of Seismic Data

Weak Seismic Signal Enhancement for Low Signal-to-Noise Ratio Data Using Adaptive Nonstationary Signal Decomposition