Article

Seismic random noise suppression based on the discrete cosine transform

Authors:
To read the full-text of this research, you can request a copy directly from the author.

Abstract

This paper proposes a seismic random noise suppression method using the predictive filter in the discrete cosine transform (DCT) domain, and then evaluates its performance. In comparison with the discrete Fourier transform (DFT), DCT can represent seismic signals with fewer coefficients, i.e., DCT has superior energy compaction for seismic signal. That is why DCT can separate seismic signals from random noise better. The results from both synthetic and real data show that the proposed method achieves better performance for random noise suppression and signal preservation compared with the f-x prediction filter.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the author.

... Signal-to-noise ratio, Resolution, and Fidelity are the three major evaluation criteria for seismic data quality, among which high signal-to-noise ratio is the basis of those (high signal-to-noise ratio, high resolution, and high fidelity) [1] . Methods include: Discrete cosine transform [2] , F-K [3] , Wavelet transform, Radon transform. The denoising based on compressed sensing is to use the corresponding sparse matrix in the sparse domain to transform the actual data into a coefficient matrix, and then reconstruct the seismic trace through the coefficient matrix, and then process the coefficient matrix to finally achieve the desired denoising effect. ...
... The principle of the orthogonal matching tracking algorithm is based on the principle of the MP algorithm. It uses the atoms with the greatest correlation with the error to be selected, and the linear combination of atoms is used to reconstruct the optimal approximate signal [2] ; and the specific flow of the OMP algorithm is shown in the figure. 2 shown: ...
Article
Full-text available
The quality of seismic data has always been an important factor restricting the results of data processing and interpretation. With the development of new technologies, the amount of collected data is also increasing, and the existence of noise is inevitable. Therefore, denoising has become a major problem. Therefore, the compressed sensing theory is used in the article, and the K-SVD learning dictionary is used as a sparse base to perform denoising tests on the model data of different layers, and then the actual seismic profile of a certain work area is processed. The results show that the K-SVD learning dictionary can better denoise the signal and reconstruct the original signal, which lays a good foundation for the subsequent interpretation work.
... In addition, many time-frequency tools have been applied to suppress the random noise in seismic data. The Fourier transform is the basis of many methods of seismic data denoising [15], including a discrete cosine transform (DCT) [16] and domain deconvolution filtering [17]. In order to make the spectrum effectively represent the local features of time signals, the method of time-windowing is proposed. ...
Article
Deep learning has been successfully applied to image denoising. In this study, we take one step forward by using deep learning to suppress random noise in poststack seismic data from the aspects of network architecture and training samples. On the one hand, poststack seismic data denoising mainly aims at 3-D seismic data. We designed an end-to-end 3-D denoising convolutional neural network (3-D-DnCNN) that takes raw 3-D cubes as input in order to better extract the features of the 3-D spatial structure of poststack seismic data. On the other hand, denoising images with deep learning require noisy-clean sample pairs for training. In the field of seismic data processing, researchers usually try their best to suppress noise by using complex processes that combine different methods, but clean labels of seismic data are not available. In addition, building training samples in field seismic data has become an interesting but challenging problem. Therefore, we propose a training sample selection method that contains a complex workflow to produce comparatively ideal training samples. Experiments in this study demonstrate that deep learning can directly learn the ability to denoise field seismic data from selected samples. Although the building of the training samples may occur through a complex process, the experimental results of synthetic seismic data and field seismic data show that the 3-D-DnCNN has learned the ability to suppress the Gaussian noise and super-Gaussian noise from different training samples. Moreover, the 3-D-DnCNN network has better denoising performance toward arc-like imaging noise. In addition, we adopt residual learning and batch normalization in order to accelerate the training speed. After network training is satisfactorily completed, its processing efficiency can be significantly higher than that of conventional denoising methods.
Article
In 3C or high density seismic surveys, point receiver, viz. one sensor at a station, is often adopted. However records acquired by this approach are always dominated by background noise. This kind of noise cannot be removed substantially by conventional noise attenuation methods. We propose in this paper a new method for suppressing the background noise on point receiver data. The method includes two steps. Firstly, we identify noisy traces based on the energy distribution before the first break. Then we interpolate the amplitude spectrum envelop of the noisy trace based on the amplitude spectrum envelop of an adjacent normal trace in frequency domain. Therefore noise on the noisy trace is minimized as its spectrum envelops of noisy trace is restricted. Not only data S/N ratio is improved but also the amplitude is preserved because the suppression is done only on noisy traces and suppression frequency is out of the dominant frequency. The proposed method has been applied to field seismic data and good result has been achieved. Noise from producing wells and vehicles has been effectively suppressed and data quality of stack sections is greatly improved.
Article
Based on a good relevance of the seismic data characteristics, singular value decomposition (SVD) technique can separate signals from noise. The seismic data can be decomposed by SVD, and then divided into a number of singular values in accordance with their energy values. The intrinsic images with big singular values can be mainly viewed as signals. We select the intrinsic images with the biggest singular values to reconstruct seismic records. This can effectively remove the random noise of seismic data. Post-stack seismic data can be obtained after a few processes such as static corrections, amplitude compensation, deconvolution, NMO and stack provided by GeoEast processing sub-system. Processing data results after forward modeling and Re-ou 3D in Liaohe Basin show that SVD can both effectively remove noise from seismic data and enhance seismic data signal to noise ratio.
Article
Real seismic signals can usually be represented as a linear combination of multiple basic waveforms of different morphological characteristics, such as powerline single frequency interference noise and effective seismic signal, ground roll and body wave signals. By selecting a single mathematical transformation method, it will hard to achieve sparse representations of seismic signals. Morphological component analysis (MCA) theory has been developed in recent years, which can sparsely decompose complex signals through the augmented dictionary of basic mathematical transformations. According to the waveform divergence of ground-roll and body wave signals, this paper proposes a ground-roll attenuation method based on the morphological component analysis theory. To match with the requirements of the MCA, the 1D stationary wavelet transform (SWT) is chosen as the sparse representation dictionary of ground-roll due to its low frequency and narrow spectral bandwidth nature. Meanwhile, the local discrete cosine transform (LDCT) is chosen as the sparse representation dictionary of body waves due to its local interdependency characteristics. The optimization model on basis of the MCA is then built on the two amalgamated dictionaries and properly solved to obtain the final signal-noise separation results. The theoretical and real data processing results confirm that the method can not only attenuate strong ground-roll noise in seismic records but also preserves the waveform characteristics and spectral bandwidth of effective signal well. The method can provide high quality data for subsequent processing and analysis.
Article
Common-reflection-point (CRP) gather is a bridge that connects seismic data and petrophysical parameters. Pre-stack attributes extraction and pre-stack inversion, both of them are important means of reservoir prediction. Quality of CRP gather usually has great impact on the accuracy of seismic exploration. Therefore, pre-stack CRP gathers noise suppression technology becomes a major research direction. Based on the vector decomposition principle, here we propose a method to suppress noise. This method estimates optimal unit vectors by searching in various directions and then suppresses noise through vector angle smoothing and restriction. Model tests indicate that the proposed method can separate effective signal from noise very well and suppress random noise effectively in single wavenumber case. Application of our method to real data shows that the method can recover effective signal with good amplitude preserved from pre-stack noisy seismic data even in the case of low signal to noise ratio (SNR). © 2015, China University of Geosciences and Springer-Verlag Berlin Heidelberg.
Article
Full-text available
The separation of signal and noise is a central issue in seismic data processing. The noise is both random and coherent in nature, the coherent part often masquerading as signal. In this tutorial, we present some approaches to signal isolation, in which stacking is a central concept. Our methodology is to transform the data to a domain where noise and signal are separable, a goal that we attain by means of inversion. We illustrate our ideas with some of our favorite transformations: wavelets, eigenvectors, and Radon transforms. We end with the notion of risk, baseball, and the Stein estimator.
Article
The ordinary time or space domain Wiener filter is conventionally obtained for real‐valued inputs and real‐valued desired outputs. Situations exist for which these variables are complex valued. It is possible to develop a solution for the resulting complex Wiener filter in terms of complex autocorrelation and crosscorrelation functions. An algorithm for the efficient evaluation of the complex filter weights is also available.