Self-training and learning the waveform features of microseismic data using an adaptive dictionary

Geophysics ◽  
2020 ◽  
Vol 85 (3) ◽  
pp. KS51-KS61 ◽  
Author(s):  
Hang Wang ◽  
Quan Zhang ◽  
Guoyin Zhang ◽  
Jinwei Fang ◽  
Yangkang Chen
Keyword(s):  
Dictionary Learning ◽  
Single Channel ◽  
Signal To Noise Ratio ◽  
Random Noise ◽  
Microseismic Monitoring ◽  
Training Data ◽  
Learning Framework ◽  
Microseismic Data ◽  
Useful Signal ◽  
Fracturing Process

Microseismic monitoring is an indispensable technique in characterizing the physical processes that are caused by extraction or injection of fluids during the hydraulic fracturing process. Microseismic data, however, are often contaminated with strong random noise and have a low signal-to-noise ratio (S/N). The low S/N in most microseismic data severely affects the accuracy and reliability of the source localization and source-mechanism inversion results. We have developed a new denoising framework to enhance the quality of microseismic data. We use the method of adaptive sparse dictionaries to learn the waveform features of the microseismic data by iteratively updating the dictionary atoms and sparse coefficients in an unsupervised way. Unlike most existing dictionary learning applications in the seismic community, we learn the features from 1D microseismic data, thereby to learn 1D features of the waveforms. We develop a sparse dictionary learning framework and then prepare the training patches and implement the algorithm to obtain favorable denoising performance. We use extensive numerical examples and real microseismic data examples to demonstrate the validity of our method. Results show that the features of microseismic waveforms can be learned to distinguish signal patches and noise patches even from a single channel of microseismic data. However, more training data can make the learned features smoother and better at representing useful signal components.

scholarly journals Simultaneous denoising of multicomponent microseismic data by joint sparse representation with dictionary learning

Geophysics ◽  
2019 ◽  
Vol 84 (5) ◽  
pp. KS155-KS172
Author(s):  
Jie Shao ◽  
Yibo Wang ◽  
Yi Yao ◽  
Shaojiang Wu ◽  
Qingfeng Xue ◽  
...  
Keyword(s):  
Sparse Representation ◽  
Dictionary Learning ◽  
Matching Pursuit ◽  
Signal To Noise Ratio ◽  
Random Noise ◽  
Time Consistency ◽  
Denoising Method ◽  
Microseismic Data ◽  
Joint Sparse Representation ◽  
Consistency Constraint

Microseismic data usually have a low signal-to-noise ratio, necessitating the application of an effective denoising method. Most conventional denoising methods treat each component of multicomponent data separately, e.g., denoising methods with sparse representation. However, microseismic data are often acquired with a 3C receiver, especially in borehole monitoring cases. Independent denoising ignores the relative amplitudes and vector relationships between different components. We have developed a new simultaneous denoising method for 3C microseismic data based on joint sparse representation. The three components are represented by different dictionary atoms; the dictionary can be fixed or adaptive depending on the dictionary learning method that is used. Our method adds an extra time consistency constraint with simultaneous transformation of 3C data. The joint sparse optimization problem is solved using the extended orthogonal matching pursuit. Synthetic microseismic data with a double-couple source mechanism and two field downhole microseismic data were used for testing. Independent denoising of 1C data with the fixed dictionary method and simultaneous denoising of 3C data with the fixed dictionary and dictionary learning (3C-DL) methods were compared. The results indicate that among the three methods, the 3C-DL method is the most effective in suppressing random noise, preserving weak signals, and restoring polarization information; this is achieved by combining the time consistency constraint and dictionary learning.

scholarly journals Three-Component Microseismic Data Denoising Based on Re-Constrain Variational Mode Decomposition

2021 ◽  
Vol 11 (22) ◽  
pp. 10943
Author(s):  
Zhili Chen ◽  
Peng Wang ◽  
Zhixian Gui ◽  
Qinghui Mao
Keyword(s):  
Signal To Noise Ratio ◽  
Solution Process ◽  
Vertical Component ◽  
Microseismic Monitoring ◽  
Center Frequency ◽  
Variational Mode Decomposition ◽  
Denoising Method ◽  
Microseismic Data ◽  
Mode Decomposition ◽  
Processing Step

Microseismic monitoring is an important technology used to evaluate hydraulic fracturing, and denoising is a crucial processing step. Analyses of the characteristics of acquired three-component microseismic data have indicated that the vertical component has a higher signal-to-noise ratio (SNR) than the two horizontal components. Therefore, we propose a new denoising method for three-component microseismic data using re-constrain variational mode decomposition (VMD). In this method, it is assumed that there is a linear relationship between the modes with the same center frequency among the VMD results of the three-component data. Then, the decomposition result of the vertical component is used as a constraint to the whole denoising effect of the three-component data. On the basis of VMD, we add a constraint condition to form the re-constrain VMD, and deduce the corresponding solution process. According to the synthesis data analysis, the proposed method can not only improve the SNR level of three-component records, it also improves the accuracy of polarization analysis. The proposed method also achieved a satisfactory effect for field data.

scholarly journals A Natural Images Pre-Trained Deep Learning Method for Seismic Random Noise Attenuation

2022 ◽  
Vol 14 (2) ◽  
pp. 263
Author(s):  
Haixia Zhao ◽  
Tingting Bai ◽  
Zhiqiang Wang
Keyword(s):  
Deep Learning ◽  
Seismic Data ◽  
Field Data ◽  
Noise Suppression ◽  
Signal To Noise Ratio ◽  
Random Noise ◽  
Natural Images ◽  
Training Data ◽  
Learning Approaches ◽  
Learning Method

Seismic field data are usually contaminated by random or complex noise, which seriously affect the quality of seismic data contaminating seismic imaging and seismic interpretation. Improving the signal-to-noise ratio (SNR) of seismic data has always been a key step in seismic data processing. Deep learning approaches have been successfully applied to suppress seismic random noise. The training examples are essential in deep learning methods, especially for the geophysical problems, where the complete training data are not easy to be acquired due to high cost of acquisition. In this work, we propose a natural images pre-trained deep learning method to suppress seismic random noise through insight of the transfer learning. Our network contains pre-trained and post-trained networks: the former is trained by natural images to obtain the preliminary denoising results, while the latter is trained by a small amount of seismic images to fine-tune the denoising effects by semi-supervised learning to enhance the continuity of geological structures. The results of four types of synthetic seismic data and six field data demonstrate that our network has great performance in seismic random noise suppression in terms of both quantitative metrics and intuitive effects.

Dictionary learning and shift-invariant sparse coding denoising for CSEM data combined with CEEMD

Geophysics ◽  
2021 ◽  
pp. 1-52
Author(s):  
Guang Li ◽  
Zhushi He ◽  
Jing Tian Tang ◽  
Juzhi Deng ◽  
Xiaoqiong Liu ◽  
...  
Keyword(s):  
Dictionary Learning ◽  
Sparse Coding ◽  
Learning Algorithm ◽  
Signal To Noise Ratio ◽  
Estimation Method ◽  
Real Data ◽  
Ensemble Empirical Mode Decomposition ◽  
Baseline Drift ◽  
Mode Decomposition ◽  
Useful Signal

Controlled-source electromagnetic (CSEM) data recorded in industrialized areas are inevitably contaminated by strong cultural noise. Traditional noise attenuation methods are often ineffective for intricate aperiodic noise. To address the problem mentioned above, we propose a novel noise isolation method based on fast Fourier transform (FFT), complementary ensemble empirical mode decomposition (CEEMD) and shift-invariant sparse coding (SISC, an unsupervised machine learning algorithm under a data-driven framework). First, large powerline noise is accurately subtracted in the frequency domain. Then, the CEEMD based algorithm is used to correct the large baseline drift. Finally, taking advantage of the sparsity of periodic signals, SISC is applied to autonomously learn a feature atom (the useful signal with a length of one period) from the detrended signal and recover CSEM signal with high accuracy. We demonstrate the performance of the SISC by comparing with other three promising signal processing methods, including the mathematic morphology filtering (MMF), soft-threshold wavelet filtering, and K-SVD (another dictionary learning method) sparse decomposition. Experimental results illustrate that SISC provides the best performance. Robustness test results show that SISC can increase the signal-to-noise ratio (SNR) of noisy signal from 0 dB to more than 15 dB. Case studies of synthetic and real data collected in the Chinese Provinces Sichuan and Yunnan show that the proposed method is capable of effectively recovering the useful signal from the observed data contaminated with different kinds of strong ambient noise. The curves of U/I and apparent resistivity after applying the proposed method improved greatly. Moreover, the proposed method performs better than the robust estimation method based on correlation analysis.

scholarly journals A Novel Hyperspectral Endmember Extraction Algorithm Based on Online Robust Dictionary Learning

2019 ◽  
Vol 11 (15) ◽  
pp. 1792 ◽  
Author(s):  
Xiaorui Song ◽  
Lingda Wu
Keyword(s):  
Dictionary Learning ◽  
Large Scale ◽  
Hyperspectral Image ◽  
Signal To Noise Ratio ◽  
Extraction Methods ◽  
Computational Time ◽  
Endmember Extraction ◽  
Learning Framework ◽  
Extraction Algorithm ◽  
High Level

Due to the sparsity of hyperspectral images, the dictionary learning framework has been applied in hyperspectral endmember extraction. However, current endmember extraction methods based on dictionary learning are not robust enough in noisy environments. To solve this problem, this paper proposes a novel endmember extraction approach based on online robust dictionary learning, termed EEORDL. Because of the large scale of the hyperspectral image (HSI) data, an online scheme is introduced to reduce the computational time of dictionary learning. In the proposed algorithm, a new form of the objective function is introduced into the dictionary learning process to improve the robustness for noisy HSI data. The experimental results, conducted with both synthetic and real-world hyperspectral datasets, illustrate that the proposed EEORDL outperforms the state-of-the-art approaches under different signal-to-noise ratio (SNR) conditions, especially for high-level noise.

scholarly journals Single-channel speech enhancement based on joint constrained dictionary learning

2021 ◽  
Vol 2021 (1) ◽  
Author(s):  
Linhui Sun ◽  
Yunyi Bu ◽  
Pingan Li ◽  
Zihao Wu
Keyword(s):  
Speech Enhancement ◽  
Dictionary Learning ◽  
Single Channel ◽  
Signal To Noise Ratio ◽  
Subjective Evaluation ◽  
Noise Signal ◽  
Noisy Signal ◽  
Optimization Function ◽  
The Cross ◽  
Projection Error

AbstractTo improve the performance of speech enhancement in a complex noise environment, a joint constrained dictionary learning method for single-channel speech enhancement is proposed, which solves the “cross projection” problem of signals in the joint dictionary. In the method, the new optimization function not only constrains the sparse representation of the noisy signal in the joint dictionary, and controls the projection error of the speech signal and noise signal on the corresponding sub-dictionary, but also minimizes the cross projection error and the correlation between the sub-dictionaries. In addition, the adjustment factors are introduced to balance the weight of constraint terms to obtain the joint dictionary more discriminatively. When the method is applied to the single-channel speech enhancement, speech components of the noisy signal can be more projected onto the clean speech sub-dictionary of the joint dictionary without being affected by the noise sub-dictionary, which makes the quality and intelligibility of the enhanced speech higher. The experimental results verify that our algorithm has better performance than the speech enhancement algorithm based on discriminative dictionary learning under white noise and colored noise environments in time domain waveform, spectrogram, global signal-to-noise ratio, subjective evaluation of speech quality, and logarithmic spectrum distance.

scholarly journals Orthogonality-Promoting Dictionary Learning via Bayesian Inference

2019 ◽  
Vol 33 ◽  
pp. 4472-4479
Author(s):  
Lei Luo ◽  
Jie Xu ◽  
Cheng Deng ◽  
Heng Huang
Keyword(s):  
Bayesian Inference ◽  
Dictionary Learning ◽  
Expectation Maximization Algorithm ◽  
Training Data ◽  
Training Set ◽  
Learning Framework ◽  
Learning Tasks ◽  
Sparsity Level ◽  
Better Than ◽  
Non Parametric

Dictionary Learning (DL) plays a crucial role in numerous machine learning tasks. It targets at finding the dictionary over which the training set admits a maximally sparse representation. Most existing DL algorithms are based on solving an optimization problem, where the noise variance and sparsity level should be known as the prior knowledge. However, in practice applications, it is difficult to obtain these knowledge. Thus, non-parametric Bayesian DL has recently received much attention of researchers due to its adaptability and effectiveness. Although many hierarchical priors have been used to promote the sparsity of the representation in non-parametric Bayesian DL, the problem of redundancy for the dictionary is still overlooked, which greatly decreases the performance of sparse coding. To address this problem, this paper presents a novel robust dictionary learning framework via Bayesian inference. In particular, we employ the orthogonality-promoting regularization to mitigate correlations among dictionary atoms. Such a regularization, encouraging the dictionary atoms to be close to being orthogonal, can alleviate overfitting to training data and improve the discrimination of the model. Moreover, we impose Scale mixture of the Vector variate Gaussian (SMVG) distribution on the noise to capture its structure. A Regularized Expectation Maximization Algorithm is developed to estimate the posterior distribution of the representation and dictionary with orthogonality-promoting regularization. Numerical results show that our method can learn the dictionary with an accuracy better than existing methods, especially when the number of training signals is limited.

Feasibility of monitoring hydraulic fracturing using time-lapse audio-magnetotellurics

Geophysics ◽  
2012 ◽  
Vol 77 (4) ◽  
pp. WB119-WB126 ◽  
Author(s):  
Lanfang He ◽  
Xiumian Hu ◽  
Ligui Xu ◽  
Zhanxiang He ◽  
Weili Li
Keyword(s):  
Hydraulic Fracturing ◽  
Case History ◽  
Oil And Gas ◽  
Signal To Noise Ratio ◽  
Time Lapse ◽  
Microseismic Monitoring ◽  
High Signal ◽  
Hydraulic Fractures ◽  
Industry Waste ◽  
Fracturing Process

Hydraulic fracturing is widely used for initiating and subsequently propagating fractures in reservoir strata by means of a pressurized fluid to release oil and gas or to store industry waste. Downhole or surface microseismic monitoring is commonly used to characterize the hydraulically induced fractures. However, in some cases, downhole microseismic monitoring can be difficult due to the limitation imposed by boreholes. Surface microseismic monitoring often faces difficulties acquiring high signal-to-noise ratio data because of the on-site noise from hydraulic fracturing process. Research and field observations indicate that injecting conductive slurry or water into a strata may generate distinct time-lapse electromagnetic anomalies between pre- and posthydraulic fracturing. These anomalies provide a means for characterizing the hydraulic fracturing using time-lapse electromagnetic methods. We examined the time-lapse variation over an hour, one day, one month, and two years of observed audio-magnetotellurics (AMT) resistivity and the 1D and 3D AMT modeling result of the variation pre- and posthydraulic fracturing. There is also a successful case history of applying the time-lapse AMT to map hydraulic fractures. Observed data indicate that the variation of AMT resistivity is normally less than 6% apart from the data of the dead band and some noisy data. Modeling results show the variation pre- and posthydraulic fracturing is larger than 30% at the frequency point lower than 100 Hz. The case history indicates that time-lapse magnetotelluric monitoring may form a new way to characterize the hydraulic fracture.

Least-squares decomposition with time–space constraint for denoising microseismic data

2020 ◽  
Vol 222 (3) ◽  
pp. 1864-1880
Author(s):  
Yangkang Chen ◽  
Wei Chen ◽  
Yufeng Wang ◽  
Min Bai
Keyword(s):  
Least Squares ◽  
Signal To Noise Ratio ◽  
Random Noise ◽  
Microseismic Data ◽  
Smoothness Constraint ◽  
Time Space ◽  
Wide Range ◽  
The Stability ◽  
And Inversion

SUMMARY Microseismic data are usually of low signal-to-noise ratio (SNR), which makes it difficult to utilize the microseismic waveforms for imaging and inversion. We develop a useful denoising algorithm based on a non-stationary least-squares decomposition model to enhance the quality of microseismic signals. The microseismic signals are assumed to be represented by a superposition of several smoothly variable components. We construct a least-squares inverse problem to solve for the the smooth components. We constrain the least-squares inversion via both time and space constraints. The temporal smoothness constraint is applied to ensure the stability when calculating the non-stationary autoregression coefficients. The space-smoothness constraint is applied to extract the spatial correlation among multichannel microseismic traces. The new algorithm is validated via several synthetic and real microseismic data and are proved to be effective. Comparison with the state-of-the-art algorithms demonstrates that the proposed method is more powerful in suppressing random noise of a wide range of levels than its competing methods.

Signal-to-noise ratio enhancement for downhole microseismic data based on 3D shearlet transform

Geophysics ◽  
2019 ◽  
Vol 84 (3) ◽  
pp. V207-V218
Author(s):  
Juan Li ◽  
Yuan Li ◽  
Shou Ji ◽  
Yue Li ◽  
Zhihong Qian
Keyword(s):  
Noise Suppression ◽  
Signal To Noise Ratio ◽  
Random Noise ◽  
Great Difficulty ◽  
Threshold Function ◽  
Signal To Noise ◽  
Shearlet Transform ◽  
Microseismic Data ◽  
3D Matrix ◽  
Noise Ratio

Downhole microseismic data are characterized for their high frequency and small amplitude, which bring great difficulty for noise suppression. We present a random noise attenuation method for downhole microseismic data based on the 3D shearlet transform (3DST). In contrast to the 2D shearlet, 3DST takes into account the correlation among three components of downhole microseismic. With the help of correlation among the data, downhole microseismic data are reassembled into a new 3D matrix and then transformed to the shearlet domain. After the analysis of the coefficients’ energy and the high-order cumulant on each scale, an efficient threshold function is proposed. We apply a small threshold to the coefficients associated with the signal’s scales, and a large threshold is chosen for the scales of the noise. Experimental results indicate that the algorithm significantly improves the signal-to-noise ratio of the microseismic data and effectively preserves a valid signal.

Sign in / Sign up

Export Citation Format

Share Document