Classification and Suppression of Blending Noise Using Convolutional Neural Networks

Baardman, Rolf; Tsingas, C.

doi:10.2118/194731-ms

Cited by 15 publications

(18 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…CNNs-based models have also proven successful at coherent noise attenuation within seismic data. For example, standard CNNs has been utilized for the likes of deblending [1,36] and linear noise removal [30]. The DnCNN architecture has also been employed for the suppression of linear noise within seismic data [46,42].…”

Section: Introductionmentioning

confidence: 99%

Coherent noise suppression via a self-supervised blind-trace deep learning scheme

Liu¹,

Birnie²,

Alkhalifah³

2022

Preprint

View full text Add to dashboard Cite

Coherent noise regularly plagues seismic recordings, causing artefacts and uncertainties in products derived from down-the-line processing and imaging tasks. The outstanding capabilities of deep learning in denoising of natural and medical images have recently spur a number of applications of neural networks in the context of seismic data denoising. A limitation of the majority of such methods is that the deep learning procedure is supervised and requires clean (noise-free) data as a target for training the network. Blindspot networks were recently proposed to overcome this requirement, allowing training to be performed directly on the noisy field data as a powerful suppressor of random noise. A careful adaptation of the blind-spot methodology allows for an extension to coherent noise suppression. In this work, we expand the methodology of blind-spot networks to create a blind-trace network that successfully removes trace-wise coherent noise. Through an extensive synthetic analysis, we illustrate the denoising procedure's robustness to varying noise levels, as well as varying numbers of noisy traces within shot gathers. It is shown that the network can accurately learn to suppress the noise when up to 60% of the original traces are noisy. Furthermore, the proposed procedure is implemented on the Stratton 3D field dataset and is shown to restore the previously corrupted direct arrivals. Our adaptation of the blind-spot network for self-supervised, trace-wise noise suppression could lead to other use-cases such as the suppression of coherent noise arising from wellsite activity, passing vessels or nearby industrial activity.Preprint. Under review.

show abstract

Section: Introductionmentioning

confidence: 99%

Coherent noise suppression via a self-supervised blind-trace deep learning scheme

Liu¹,

Birnie²,

Alkhalifah³

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…In addition, robust implementations of rank reduction filtering have been widely applied as a deblending tool that effectively removes blending noise, which is manifested as a non-Gaussian erratic noise, or as random noise (Wason et al 2014;Sacchi 2015, 2016;Maraschini et al 2016;Zu et al, 2017;Jeong et al 2020a). Recently, there have been numerical studies that suggest novel algorithms to seismic deblending based on sparse coding and dictionary learning (Zu et al 2019), convolutional neural networks (Baardman and Tsingas 2019;Baardman and Hegge 2020;Sun et al 2020) and unsupervised anomaly detection (Jeong et al 2020b). Alternatively, another category was introduced by an inversionbased deblending framework that iteratively predicts and subtracts blending noise from the blended data (Abma and Yan 2009;Abma et al 2010;Mahdad et al 2011).…”

Section: Introductionmentioning

confidence: 99%

Deblending and merging of 3D multi‐sweep seismic blended data

Jeong

Tsingas

Almubarak

2021

Geophysical Prospecting

View full text Add to dashboard Cite

Seismic blended source acquisition, also referred to as simultaneous source acquisition, is a cost-effective technology that achieves a significant reduction in acquisition cycle time and increases seismic crew field productivity. The dispersed source array is a blended acquisition field technique that simultaneously employs sources emitting different types of sweeps (i.e. multi-sweep), in terms of frequency bandwidth and length, which ultimately will result in a full broadband seismic data. In this paper, deblending of 3D multi-sweep seismic blended data and the subsequent merging of the data volumes having different frequency bandwidths will be discussed. In specific data domains where the signal component is coherent, interference shots (i.e. blending noise) are randomly distributed in the data space according to its own shot firing time. Therefore, the deblending process, which separates interference shots from a signal component, becomes a noise attenuation problem. A sparse inversion methodology is applied in the frequency-wavenumber-wavenumber (f-k x -k y ) domain to attenuate blending noise. By applying this deblending methodology to both dispersed source array's low-and mid-high-frequency bandwidths, we obtained high-quality deblending results. For both frequency bandwidths of the deblended dispersed source array data, additional effort was made to combine the two datasets to a single broadband data volume. Consequently, deblending and merging of the dispersed source array blended data generated a broadband, deblended and well-balanced seismic volume suitable for further processing and reservoir characterization applications.

show abstract

“…Xiong et al (2018) trained a CNN to automatically detect and map fault zones using 3D seismic images, whereas Wu et al (2019) proposed to use CNNs to pick the first arrivals of microseismic events. Baardman et al (2018Baardman et al ( , 2019 proposed the use of a CNN to classify data patches in a "blended" and "non-blended" class. A second, regression based, CNN was then employed to deblend the "blended" patches, but only synthetic data were considered.…”

Section: Introductionmentioning

confidence: 99%

A convolutional neural network approach to deblending seismic data

et al. 2020

View full text Add to dashboard Cite

For economic and efficiency reasons, blended acquisition of seismic data is becoming increasingly commonplace. Seismic deblending methods are computationally demanding and normally consist of multiple processing steps. Furthermore, the process of selecting parameters is not always trivial. Machine-learning-based processing has the potential to significantly reduce processing time and to change the way seismic deblending is carried out. We have developed a data-driven deep-learning-based method for fast and efficient seismic deblending. The blended data are sorted from the common-source to the common-channel domain to transform the character of the blending noise from coherent events to incoherent contributions. A convolutional neural network is designed according to the special characteristics of seismic data and performs deblending with results comparable to those obtained with conventional industry deblending algorithms. To ensure authenticity, the blending was performed numerically and only field seismic data were used, including more than 20,000 training examples. After training and validating the network, seismic deblending can be performed in near real time. Experiments also indicate that the initial signal-to-noise ratio is the major factor controlling the quality of the final deblended result. The network is also demonstrated to be robust and adaptive by using the trained model to first deblend a new data set from a different geologic area with a slightly different delay time setting and second to deblend shots with blending noise in the top part of the record.

show abstract

Classification and Suppression of Blending Noise Using Convolutional Neural Networks

Cited by 15 publications

References 3 publications

Coherent noise suppression via a self-supervised blind-trace deep learning scheme

Coherent noise suppression via a self-supervised blind-trace deep learning scheme

Deblending and merging of 3D multi‐sweep seismic blended data

A convolutional neural network approach to deblending seismic data

Contact Info

Product

Resources

About