 Research
 Open access
 Published:
SeisDeNet: an intelligent seismic data Denoising network for the internet of things
Journal of Cloud Computing volume 12, Article number: 34 (2023)
Abstract
Deep learning (DL) has attracted tremendous interest in various fields in last few years. Convolutional neural networks (CNNs) based DL architectures have been successfully applied in computer vision, medical image processing, remote sensing, and many other fields. A recent work has proved that CNNs based models can also be used to handle geophysical problems. Due to noises in seismic signals acquired by geophone equipment this kind of important multimedia resources cannot be effectively utilized in practice. To this end, from the perspective of seismic exploration informatization, this paper takes informatization data in seismic signal acquisition and energy exploration field using cuttingedge technologies such as Internet of things and cloud computing as the research object, presenting a novel CNNs based seismic data denoising (SeisDeNet) architecture is suggested. Firstly, a multiscale residual dense (MSRD) block is built to leverage the characteristics of seismic data. Then, a deep MSRD network (MSRDN) is proposed to restore the noisy seismic data in a coarsetofine manner by using cascading MSRDs. Additionally, the denoising problem is formulated into predicting transformdomain coefficients, by which noises can be further removed by MSRDNs while richer structure details are preserved comparing with the results in spatial domain. By using synthetic seismic records, public SEG and EAGE salt and overthrust seismic model and real field seismic data, the proposed method is qualitatively and quantitatively compared with other leading edge schemes to evaluate it performance, and some results shows that the proposed scheme can produce data with higher quality evaluation while maintaining far more useful data comparing with other schemes. The feasibility of this approach is confirmed by the denoising results, and this approach is shown to be promising in suppressing the seismic noise automatically.
Introduction
With the construction of the national digital detection network, the observation methods and technologies of the new Internet of things are gradually applied [1,2,3,4], which basically realizes the digitization, networking and integration of seismic monitoring and exploration, and the ability of seismic monitoring and exploration has been greatly improved. The application of digital technology provides rich firsthand data for seismic scientific research and lays a good foundation for explanations and predictions of seismic exploration. Noise attenuation is crucial to obtain highquality data from the seismic signals which are collected with geophone sensing equipment and networks [4]. But this usually badly interfered with and distorted by noises in seismic exploration. It has an important impact on oil and gas exploration.
In order to well denoise the big data generated by the Internet of things, many seismic denoising methods have been proposed. Among previously proposed seismic denoising methods [5,6,7,8,9,10,11,12,13], the most effective ones include dictionary learning schemes [12, 13] and sparse transform schemes [6, 7, 9,10,11] Convolutional neural networks (CNNs) based methods [14,15,16,17,18,19,20,21,22] show powerful learning ability and have been intensively applied in various fields, i.e., computer vision tasks including nature image superresolution (SR), finegrained visual categorization [23, 24], deep optimized MRI technique [25] for clinical applications. Even more, deep learning has been used to solve datacenterscale traffic optimization, with outstanding performance. Aiming to fulfill high precision seismic exploration and appropriate seismic signal denoising CNN model is established in the present work. The advanced deep learning can promote the development of geophysical technology. And vice versa, task requirements of geophysics can accelerate the progress of artificial intelligence technology.
To remarkably enhance signal to noise ratios (SNR) and adapt to high precision seismic exploration, a novel transform domain based CNN architecture for denoising seismic data generated by the Internet of Things is presented in this paper, and primary contributions include:

(i)
To fully exploit features from seismic data, MSRD block is proposed and built for restoring seismic data with noises.

(ii)
Denoising problem is formulated into predicting transformdomain coefficients, by which noises can be further removed with MSRDN while more detail data is preserved comparing with the results in spatial domain.

(iii)
Our method is qualitatively and quantitatively evaluated with synthetic seismic records, public SEG and EAGE salt and overthrust seismic model and real field seismic data. The proposed method can get signals with higher PSNRs and far more useful data comparing with other leading edge ones. Essentially, it shows more considerable denoising performance under higher noise level.
The reminder of the manuscript is structured as the following: Section 2 presents a brief review of relate studies. Section 3 proposes a novel scheme, and it is validated in Section 4. Lastly, Section 5 summarizes the work.
Related work
For seismic exploration data from the Internet of things, many seismic denoising methods have been developed so far. The seismic denoising methods were proposed by Canales to suppress random noises and achieved a potential result [5]. Since then, several effective schemes for suppressing random noises have been proposed, such as sparse transform based methods, dictionary learning based procedures, and nonlocal means algorithm [8]. Among them, sparse representation of seismic signals has become popular.. Conventionally, almost all denoising is performed in a transform domain [6, 7, 9,10,11]. For learning based methods [12, 13], an over complete dictionary, which is generally written as an explicit matrix, can be inferred from a series of examples, and the matrix should be trained before it is adapted to examples.
Recently, with the continuous expansion of Internet of things data and cloud platform data, there are differences in data processing. Deep learning (DL) becomes very attractive and demonstrates excellent ability in many areas, such as multimedia and machine learning, since it overcomes the shortcomings of common learning based schemes by leveraging convolutional neural network (CNN) architecture. Since the CNNbased SRCNN was first proposed in [14] for lowlevel vision, lots of methods [14,15,16,17,18,19,20,21,22] with different network depth, i.e., the deeper layer RCAN [22], have been developed. The depth of networks has a positive impact on the SR performance of these networks, which was proved by the increasing number of convolutional layers from 3 to 400. In some recent CNNbased SR, several identical feature extraction modules are connected to build the entire network [19,20,21,22], and each module is essential in this structure. Obviously, these CNNbased methods are superior over the traditional methods. Some diversified SR network structures are also proposed. Recently, attentionbased SR network models have been proposed, which uses addition operation in the output layer, avoiding the large amount of computing power consumed by convolution kernel multiplication, so as to efficiently complete the image SR.
These SRs are performed in spatial domain. But SR in transform domain can better preserve the context and texture information of an image in various layers. So, a deep wavelet SR network was proposed by Guo et al. [26] for acquiring HR images, whereas “missing details” of wavelet coefficients of LR images were predicted. Then, an orthogonally regularized deep network was suggested by the same team [27], whereas discrete cosine transformation (DCT) was integrated into CNNs. Besides, a face SR was constructed on the basis of wavelet transform and CNNs to capture local textural details and global topology information of faces [28].
Therefore, a novel transform domain based MSRDN architecture for seismic signal denoising is proposed and the method is detailed in Section 3.
Proposed method
In this paper, we propose and built a MSRD block to fully exploit features from seismic data for restoring seismic data with noises. Meanwhile, denoising problem is formulated into predicting transformdomain coefficients, by which noises can be further removed with MSRDN. First, the proposed method for seismic data denoising is outlined. Then, the architecture of the proposed MSRDN and MSRD block are briefly described. Lastly, transform domain is introduced.
Overview
Figure 1(a) presents a flowchart of our method. The noisy seismic data is first upsampled to the target highresolution size to produce resized noisy seismic data, from which one lowfrequency (LF) subband and several highfrequency (HF) subbands can be obtained by using specific transform. To predict the transform coefficients of the target clear seismic data, two deep residual networks are applied on the top of one LF subband and four HF subbands for preserving the global topology information and collecting the structure and texture information, respectively. Lastly, the target clear seismic data can be obtained via an inverse transform.
MSRD network structure for seismic Denoising
We aims to restore a clear seismic data I^{Clear} from a noisy seismic data I^{Noisy}. As shown in Fig. 1(b), our MSRDN has two parts: the shallow (SSFE) and the deep seismic signal feature extraction (DSFE) module. We solve the problem for noised and clean signals:
where θ = {W^{1}, W^{2}, …, W^{p}, b^{1}, b^{2}, …, b^{p}} is the weights and bias of our p convolutional layers. I^{Noisy} and I^{Clear} denote noised and clean signals, respective. L^{DE} is the loss function for minimizing the difference between the noised and the clean data. And N is the number of training samples.
Mean square error (MSE) function is the most popular objective optimization function in image SR [15, 17, 18], whereas training with MSE loss maybe not the best option according to Lim et al. [29]. Mean absolute error (MAE) function L^{DE} is used to reduce computations and avoid unnecessary training tricks in this work, which is defined by
Particularly, the shallow feature F_{0} is obtained from the noisy seismic data with double convolution layers as follows
where H_{SSFE1} and H_{SSFE2} are the convolution operations of these two layers. Then, the extracted F_{0} will be utilized in DSFE module. A DSFE module is composed of cascading MSRD blocks and each one can collect as much data as possible and extract more useful data. Then a 1 × 1 convolutional layer is employed to appropriately manipulate the output signals. This procedure can be expressed as
where F_{i} (i = 1, 2, …, D) is feature maps generated by MSRD blocks. [F_{1}, F_{2}, …, F_{D}] is the concatenation of the feature maps. H_{GFF} is a composite function of 1 × 1 convolutional layer. Then, feature maps I_{Output} can be obtained with global residual learning,
In the proposed DMSRDN, except the 128filter convolutional layer in Eq. 4 s, each convolutional layer has c filters, where c is set as 32, 64 and 96.
Multiscale residual dense (MSRD) block
Figure 2 shows the proposed MSRD block. Each MSRD block has mpath, which can be used for exploiting different scale characteristics features. Different from the RDN model [21], multibypass network is constructed in each module, with different convolutional kernels for different bypasses. So, the proposed model can adaptively measure mpath characteristics at various scales. This can be view as a wide and deep neural network model.
Supposing the input and output of the dth MSRD block are F_{d − 1} and F_{d}. \({F}_{d,c}^1\) \({F}_{d,c}^2\), we have
where \({Y}_{3\times 3}^i\) and \({Y}_{5\times 5}^i\) refer to the function of 3 × 3 and 5 × 5 Conv layer in the ith convolutional layer respectively, i = 1, …, c, …, C. By combining the previous information with the current multiscale information, we have retained short path information.
where \({H}_{LFF}^d\) is the composite function of the 1 × 1 convolutional layer in the dth MSRD block. σ is the ReLU function. [⋅] denotes the concatenation of feature maps by various convolutional kernels. Finally, the input information and combined multiscale information are aggregated as follows:
where F_{d} is the output of the dth MSRD block.
Transformdomain analysis for seismic data
Since wavelet can sparsely represent onedimensional signals without point discontinuities, it has been successfully used in representing digital signals. But, image functions with curves and straight lines in higher dimensions cannot be “optimally” represented with wavelet analysis [30]. Subsequently, some sparser transform methods [31,32,33] have been presented such as curvelet transform, contourlet transform (CT), nonsubsampled CT (NSCT), shearlet transform (ST), nonsubsampled ST (NSST) and compactly supported ST (CSST), etc., in which the anisotropic regularity of a surface following edges can be exploited. Wherein, CSST is optimally sparse.
Generally, a sparse representation of signals can benefit signal processing tasks. To improve the representation sparsity of signals CT was developed by Do and Vetterli [31], which has two primary features of directionality and anisotropy and be superior to curvelets, bandelets and other geometricallydriven representations in its partially easy and efficient waveletlike implementation using iterative filter banks.
Next, their sparsity is analyzed. The denoising effect is determined by the degree of representing decomposed effective signals [32]. The denoising performance becomes better and better as the sparsity of the method increases. Figure 3 shows the reconstruction errors in wavelet transform, curvelet transform, NSST and CSST domains and the data used are presented in Fig. 4(a). Clearly, NSST and CSST have the smallest approximation errors while retaining the same percentage coefficients, and the errors are close to zero when only 6% coefficients are retained, indicating its optimal sparsity. In addition, the literature [33] also indicates that compactly supported shearlets are optimally sparse.
The HF coefficients of CSST, NSST and WT are compared in Fig. 4. Obviously, CSST displays more accurate results of the curvature. We mention that these transforms can be applied in various denoising networks, with improved performance. Further experiment is conducted to evaluate the role of transform, as shown in Section 5.
Experimental results
Experiments are conducted to qualitatively and quantitatively evaluate the performance of MSRDN, with the following contrasting seismic denoising methods: traditional methods—waveletbased methods and curveletbased methods and DL based methods—VDSR [15], multiscale residual network (MSRN) [20] and residual dense network (RDN) [21].
Seismic datasets
The basic data is synthesized with lots of seismic records, such as linear, curvilinear; various dip angle and fault events, with 1000 Hz sampling frequency and 150 traces. Ricker wavelet with the following expression is used as seismic wavelet:
where t is time and f is sampling frequency. Figure 5(a) shows partial synthetic seismic data. Besides, immigrated stack profile measured with the SEG/EAGE salt and overthrust model [34] is presented in Fig. 5(b). In addition, these seismic records are rotated by 45°, 90°, 135°, 180°, 270°, and 360°, respectively, following [17, 18]. Then, random noises of various levels are added into original and rotated datasets to obtain additional expanded versions, of which 80% are selected for training and 20% for testing.
Implementation details
Our MSRDN contains 10 MSRD blocks. For training four HF subbands and one LF subband are produced by passing training seismic data with 1level NSST, and then cropped into 48 × 48 patches with an overlap of 24 pixels. The initial learning rate is 10^{− 4} for all layers and it haves every 50 epochs. The batch size is set as 64. The implementation of our method is realized under the Torch7 framework on an NVIDIA Tesla P100. The ADAM optimizer is used for updating. The approximate time of training our model is 6 hours for 200 epochs.
Comparison with traditional and leading edge methods
The denoising performance of our method is checked on synthetic seismic data in this section. Peak signaltonoise ratio (PSNR) [35] is employed to quantitatively justify the reconstruction results:
where X^{′} and X are the M × N denoised and clear seismic data, respectively; MAX_{I} is the largest possible pixel intensity value.
The comparison is conducted with an identical training set for all models, and the released codes are used for the contrast models. Tables 1 and 2 present PSNR (dB) values for comparison with bold optimal values. PSNRs of our method are significantly higher than that of other schemes when evaluated on seismic data. Besides, Figs. 6 and 7 indicate that our method presents better qualitative results, with less residual coherent and incoherent noise.
In addition, noisy seismic data of Liaohe depression, China, which is acquired in the identical data area with same excitation and reception for validating the processing result of our method, are selected the field data examples. To guarantee no valid information loss, these data are roughly processed to generate the targeted clear data by using traditional random denoising modular of large processing system. The random noises of various levels are added into the targeted data for deep learning to learn and recognize noises and effective signals. For the same reason, the real seismic data are rotated by 45°, 90°, 135°, 180°, 270°, and 360°, to obtain expanded versions. 80% of versions are used for training; the remainders are for testing. As shown in Fig. 8, Fig. 8(a) is an original noisy data, and Fig. 8(b) presents the denoised data with the proposed scheme showing some highlighted effective signals, especially in the red rectangle area, a clearer interlayer structure and enhanced continuity of the events. Overall, our proposed method can also achieve a satisfactory result for real filed seismic data.
Discussion
In this section, we mainly discuss the effectiveness of transform domain. We conduct ablation investigation. Specifically, predication of CSST coefficients is introduced into denoising of seismic data and the contribution effect is evaluated. Four methods (VDSR, MSRN, RDN and our MSRDN) are selected and integrated with CSST predictions respectively. Figure 9(a) presents the PSNRs of MSRDN with and without CSST predictions across seismic data with noises at various levels. Figure 9(b) indicates the averaged PSNRs of VDSR, MSRN and RDN with noise level 0.1 from the left to the right respectively. Significant and consistent improvement are observed across all networks and benchmarks when integrated with CSST, demonstrating that CSST predictions overperform in effectiveness as compared to spatial domain.
Conclusions
We present a CNNbased seismic data denoising method in this work. To improve the seismic denoising performance an MSRDN consisting of a set of cascading MSRD blocks is proposed to exploit features of seismic data. Additionally, by applying transformdomain operator to the network structure richer detail information can be preserved in seismic signals with noises and the seismic denoising performance is improved further. The qualitative and quantitative experimental results illuminate that our method is significantly superior over other stateoftheart ones in seismic data restoration.
Availability of data and materials
Not applicable.
References
Qiu T, Zhang L, Chen N, Zhang S, Liu W, Dapeng Oliver W (2022) Born this way: a selforganizing evolution scheme with motif for internet of things robustness. IEEE/ACM Trans Networking. https://doi.org/10.1109/TNET.2022.3178408
Chen N, Qiu T, Zilong L, Dapeng Oliver W (2022) An adaptive robustness evolution algorithm with selfcompetition and its 3D deployment for internet of things. IEEE/ACM Trans Netw 30(1): 68–381
Chen N, Qiu T, Daneshmand M, Dapeng Oliver W (2021) Robust networking: dynamic topology evolution learning for internet of things. ACM Trans Sens Netw 17(3):1–23
Chen N, Qiu T, Zhou X, Li K, Atiquzzaman M (2019) An intelligent robust networking mechanism for the internet of things. IEEE Commun Mag 57(11):91–95
Canales LL (1984) Random noise reduction. In: 54th annual international meeting of SEG technical program expanded abstracts, pp 525–527
Zhang JH, Lu JM (1997) Application of wavelet transform in removing noise and improving resolution of seismic data. J Univ Petroleum 31(12):1975–1981
Neelamani R, Baumstein A, Gillard D, Hadidi M, Soroka W (2008) Coherent and random noise attenuation using the curvelet transform. Lead Edge 27:240–248
Bonar D, Sacchi M (2012) Denoising seismic data using the nonlocal means algorithm. Geophysics 77(1):A5–A8
Xu J, Wang W, Gao JH, Chen WC (2013) Monochromatic noise removal via sparsityenabled signal decomposition method. IEEE Geosci Remote Sensing Lett 10(3):533–537
Chen Y, Fomel S (2015) EMDseislet transform. In: 85th annual international meeting of SEG technical program expanded abstracts, pp 4775–4778
Liu W, Cao S, Chen Y, Zu S (2016) An effective approach to attenuate random noise based on compressive sensing and curvelet transform. J Geophys Eng 13:135–145
Beckouche S, Ma JW (2014) Simultaneous dictionary learning and denoising for seismic data. Geophysics 79(3):A27–A31
Chen YK (2017) Fast dictionary learning for noise attenuation of multidimensional seismic data. Geophys J Int 209(1):21–31
Dong C, Loy CC, He KM, Tang XO (2014) Learning a deep convolutional network for image superresolution. In: European conference on computer vision (ECCV), pp 184–199
Dong C, Loy CC, Tang XO (2016) Accelerating the superresolution convolutional neural network. In: European Conference on Computer Vision (ECCV), pp 391–407
Kim J, Lee JK, Lee KM (2016) Accurate image superresolution using very deep convolutional networks. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp 1646–1654
Tai Y, Yang J, Liu XM (2017) Image superresolution via deep recursive residual network. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp 3147–3155
Tai Y, Yang J, Liu XM, Xu CY (2017) Memnet: a persistent memory network for image restoration. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp 4539–4547
Hui Z, Wang XM, Gao XB (2018) Fast and accurate single image superresolution via information distillation network. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp 723–731
Li JC, Fang FM, Mei KF, Zhang GX (2018) Multiscale residual network for image superresolution. In: European Conference on Computer Vision (ECCV), pp 527–542
Zhang YL, Tian YP, Kong Y, Zhong BN, Fu Y (2018) Residual dense network for image superresolution. In: The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
Zhang YL, Li KP, Li K, Wang LC, Zhong BN, Fu Y (2018) Image superresolution using very deep residual channel attention networks. European conference on computer vision (ECCV). Springer, Munich, pp 294–310
He XT, Peng YX, Zhao JJ (2018) StackDRL: Stacked deep reinforcement learning for finegrained visual categorization. In: IJCAI, pp 741–747
Sun XX, Chen L, Yang J (2019) Learning from web data using adversarial discriminative neural networks for finegrained classification. In: AAAI
Liu RS, Zhang YX, Cheng SC et al (2019) A theoretically guaranteed deep optimization framework for robust compressive sensing MRI. In: AAAI
Guo TT, Mousavi HS, Vu TH, Monga V (2017) Deep wavelet prediction for image superresolution. In: The IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), pp 104–113
Guo, T. T., Mousavi, H. S., Monga, V.: Orthogonally regularized deep networks for image superresolution. arXiv preprint arXiv:1802.02018 (2018)
Huang HB, He R, Sun ZN, Tan TN (2017) Waveletsrnet: a waveletbased cnn for multiscale face super resolution. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp 1689–1697
Lim B, Son S, Kim H, Nah S, Lee KM (2017) Enhanced deep residual networks for single image superresolution. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshop (CVPRW), pp 136–144
Mallat S (2008) A wavelet tour of signal processing: the sparse way. Academic press
Do MN, Vetterli M (2005) The contourlet transform: an efficient directional multiresolution image representation. IEEE Trans Image Process 14(12):2091–2160
Shahdoosti HR, Khayat O (2016) Image denoising using sparse representation classification and nonsubsampled shearlet transform. SIViP 10(6):1–7
Kutyniok G, Lim WQ, Reisenhofer R (2016) ShearLab 3D: faithful digital shearlet transforms based on compactly supported shearlets. ACM Trans Math Softw 42(1):5:1–5:42
Aminzadeh F, Burkhard N, Kunz T, Nicoletis L (1995) 3D modeling project: 3rd report. Lead Edge 14:125–128
Wang Z, Bovik AC, Sheikh HR, Simoncelli E (2004) Image quality assessment: from error visibility to structural similarity. IEEE Trans Image Process 13(4):600–612
Acknowledgments
We would like to thank participants to Liaohe Oilfield Company of PetroChina for their observations.
Funding
This work was supported in part by the Basic Scientific Research Project of Liaoning Provincial Department of Education under Grant No. LJKQZ2021152; in part by the National Science Foundation of China (NSFC) under Grant No. 61602226; in part by the PhD Startup Foundation of Liaoning Technical University of China under Grant No. 18–1021; in part by the Basic Scientific Research Project of Colleges and Universities in Liaoning Province under Grant No. LJKZ0358.
Author information
Authors and Affiliations
Contributions
The authors had worked equally during all this paper’s stages. All authors read and approved the final manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare that they have no competing interests.
Additional information
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Sang, Y., Peng, Y., Lu, M. et al. SeisDeNet: an intelligent seismic data Denoising network for the internet of things. J Cloud Comp 12, 34 (2023). https://doi.org/10.1186/s13677022003783
Received:
Accepted:
Published:
DOI: https://doi.org/10.1186/s13677022003783