Seismic Data Inversion with Acquisition Adaptive CNN for Geological Forward-prospecting in Tunnels

Geophysics ◽  
2021 ◽  
pp. 1-60
Author(s):  
Yuxiao Ren ◽  
Bin Liu ◽  
Senlin Yang ◽  
Duo Li ◽  
Peng Jiang
Keyword(s):  
Velocity Distribution ◽  
Seismic Velocity ◽  
Synthetic Data ◽  
Seismic Source ◽  
Structural Similarity ◽  
Superior Performance ◽  
Observation Data ◽  
Feature Maps ◽  
Tunnel Face ◽  
Acquisition Information

Seismic forward-prospecting is essential because it can identify the velocity distribution in front of the tunnel face and provide guidance for safe excavation activities. We propose a convolutional neural network (CNN)-based method to invert forward-prospecting data recorded in tunnels for accurate and rapid estimation of seismic velocity distribution. Targeting the unusual seismic acquisition setup in tunnels, we design two separate encoders to extract features from observation data recorded on both tunnel sidewalls. Subsequently, these features are concatenated to a decoder for velocity prediction. Considering the various acquisition setups used in different tunneling projects, the deep learning inversion network must be flexible in terms of the seismic source/receiver positions for practical application. We generate two auxiliary feature maps that can be used to feed acquisition information to the proposed network. The proposed network, acquisition adaptive CNN ( A2-CNN) can be trained by defining the loss function based on the L2-norm and multiscale structural similarity (MSSIM). Compared with traditional CNNs, the proposed method shows superior performance on datasets with both fixed and random acquisition setups, and also demonstrates certain robustness when handling synthetic data with field noise. Finally, we test how the network performs when feeding the modified acquisition setup information. It turns out that the inversion result will demonstrate a shift when the provided acquisition setup information shift, which verified the validity of the network and its utilization of acquisition information.

scholarly journals Focal mechanism of seismic events with a dipolar component

1995 ◽  
Vol 38 (3-4) ◽  
Author(s):  
F. Batini ◽  
M. Caputo ◽  
R. Console
Keyword(s):  
Synthetic Data ◽  
High Standard ◽  
Seismic Source ◽  
Real Data ◽  
Computer Code ◽  
Geothermal Field ◽  
Observation Data ◽  
Amplitude Data ◽  
Double Couple ◽  
Fault Mechanism

In this paper we model the geometry of a seismic source as a dislocation occurring on an elemental flat fault in an arbitrary direction with respect to the fault plane. This implies the use of a fourth parameter in addition to the three usual ones describing a simple double couple mechanism. We applied the radiation pattern obtained from the theory to a computer code written for the inversion of the observation data (amplitudes and polarities of the first onsets recorded by a network of stations). It allows the determination of the fault mechanism gener- alized in the above mentioned way. The computer code was verified on synthetic data and then applied to real data recorded by the seismic network operated by the Ente Nazionale per l'Energia Elettrica (ENEL), monitoring the geothermal field of Larderello. The experimental data show that for some events the source mechanism exhibits a significant dipolar component. However, due to the high standard deviation of the amplitude data, F-test applied to the results of the analysis shows that only for two events the confidence level for the general- ized model exceeds 90%.

scholarly journals Fractional integration of seismic wavelets in anelastic media to recover multiscale properties of impedance discontinuities

Geophysics ◽  
2018 ◽  
Vol 83 (2) ◽  
pp. V61-V71 ◽  
Author(s):  
Stephan Ker ◽  
Yves Le Gonidec
Keyword(s):  
Wavelet Transform ◽  
Fractional Integration ◽  
Synthetic Data ◽  
Seismic Source ◽  
Seismic Attributes ◽  
Data Set ◽  
Shaping Filter ◽  
Gaussian Derivative ◽  
Intrinsic Attenuation ◽  
Seismic Wavelet

Multiscale seismic attributes based on wavelet transform properties have recently been introduced and successfully applied to identify the geometry of a complex seismic reflector in an elastic medium. We extend this quantitative approach to anelastic media where intrinsic attenuation modifies the seismic attributes and thus requires a specific processing to retrieve them properly. The method assumes an attenuation linearly dependent with the seismic wave frequency and a seismic source wavelet approximated with a Gaussian derivative function (GDF). We highlight a quasi-conservation of the Gaussian character of the wavelet during its propagation. We found that this shape can be accurately modeled by a GDF characterized by a fractional integration and a frequency shift of the seismic source, and we establish the relationship between these wavelet parameters and [Formula: see text]. Based on this seismic wavelet modeling, we design a time-varying shaping filter that enables making constant the shape of the wavelet allowing retrieval of the wavelet transform properties. Introduced with a homogeneous step-like reflector, the method is first applied on a thin-bed reflector and then on a more realistic synthetic data set based on an in situ acoustic impedance sequence and a high-resolution seismic source. The results clearly highlight the efficiency of the method in accurately restoring the multiscale seismic attributes of complex seismic reflectors in anelastic media by the use of broadband seismic sources.

High-resolution seismic velocity analysis by sign-based weighted semblance

Geophysics ◽  
2021 ◽  
pp. 1-35
Author(s):  
M. Javad Khoshnavaz
Keyword(s):  
Seismic Velocity ◽  
Resolution Enhancement ◽  
Synthetic Data ◽  
Correlation Coefficients ◽  
Velocity Model ◽  
Seismic Attenuation ◽  
Velocity Analysis ◽  
Velocity Models ◽  
Amplitude Versus Offset ◽  
Amplitude Versus

Building an accurate velocity model plays a vital role in routine seismic imaging workflows. Normal-moveout-based seismic velocity analysis is a popular method to make the velocity models. However, traditional velocity analysis methodologies are not generally capable of handling amplitude variations across moveout curves, specifically polarity reversals caused by amplitude-versus-offset anomalies. I present a normal-moveout-based velocity analysis approach that circumvents this shortcoming by modifying the conventional semblance function to include polarity and amplitude correction terms computed using correlation coefficients of seismic traces in the velocity analysis scanning window with a reference trace. Thus, the proposed workflow is suitable for any class of amplitude-versus-offset effects. The approach is demonstrated to four synthetic data examples of different conditions and a field data consisting a common-midpoint gather. Lateral resolution enhancement using the proposed workflow is evaluated by comparison between the results from the workflow and the results obtained by the application of conventional semblance and three semblance-based velocity analysis algorithms developed to circumvent the challenges associated with amplitude variations across moveout curves, caused by seismic attenuation and class II amplitude-versus-offset anomalies. According to the obtained results, the proposed workflow is superior to all the presented workflows in handling such anomalies.

Colored and linear inversions to relative acoustic impedance

Geophysics ◽  
2019 ◽  
Vol 84 (2) ◽  
pp. N15-N27 ◽  
Author(s):  
Carlos A. M. Assis ◽  
Henrique B. Santos ◽  
Jörg Schleicher
Keyword(s):  
Seismic Data ◽  
Acoustic Impedance ◽  
Computational Cost ◽  
Synthetic Data ◽  
Seismic Source ◽  
Alternative Methods ◽  
Well Log ◽  
Linear Inversion ◽  
Amplitude Inversion ◽  
Band Limited

Acoustic impedance (AI) is a widely used seismic attribute in stratigraphic interpretation. Because of the frequency-band-limited nature of seismic data, seismic amplitude inversion cannot determine AI itself, but it can only provide an estimate of its variations, the relative AI (RAI). We have revisited and compared two alternative methods to transform stacked seismic data into RAI. One is colored inversion (CI), which requires well-log information, and the other is linear inversion (LI), which requires knowledge of the seismic source wavelet. We start by formulating the two approaches in a theoretically comparable manner. This allows us to conclude that both procedures are theoretically equivalent. We proceed to check whether the use of the CI results as the initial solution for LI can improve the RAI estimation. In our experiments, combining CI and LI cannot provide superior RAI results to those produced by each approach applied individually. Then, we analyze the LI performance with two distinct solvers for the associated linear system. Moreover, we investigate the sensitivity of both methods regarding the frequency content present in synthetic data. The numerical tests using the Marmousi2 model demonstrate that the CI and LI techniques can provide an RAI estimate of similar accuracy. A field-data example confirms the analysis using synthetic-data experiments. Our investigations confirm the theoretical and practical similarities of CI and LI regardless of the numerical strategy used in LI. An important result of our tests is that an increase in the low-frequency gap in the data leads to slightly deteriorated CI quality. In this case, LI required more iterations for the conjugate-gradient least-squares solver, but the final results were not much affected. Both methodologies provided interesting RAI profiles compared with well-log data, at low computational cost and with a simple parameterization.

scholarly journals Facial Action Unit Intensity Estimation via Semantic Correspondence Learning with Dynamic Graph Convolution

2020 ◽  
Vol 34 (07) ◽  
pp. 12701-12708
Author(s):  
Yingruo Fan ◽  
Jacqueline Lam ◽  
Victor Li
Keyword(s):  
Probabilistic Models ◽  
Superior Performance ◽  
Feature Maps ◽  
Dynamic Graph ◽  
Intensity Estimation ◽  
Facial Action ◽  
Learning Framework ◽  
Benchmark Datasets ◽  
Relationship Of ◽  
Semantic Correspondence

The intensity estimation of facial action units (AUs) is challenging due to subtle changes in the person's facial appearance. Previous approaches mainly rely on probabilistic models or predefined rules for modeling co-occurrence relationships among AUs, leading to limited generalization. In contrast, we present a new learning framework that automatically learns the latent relationships of AUs via establishing semantic correspondences between feature maps. In the heatmap regression-based network, feature maps preserve rich semantic information associated with AU intensities and locations. Moreover, the AU co-occurring pattern can be reflected by activating a set of feature channels, where each channel encodes a specific visual pattern of AU. This motivates us to model the correlation among feature channels, which implicitly represents the co-occurrence relationship of AU intensity levels. Specifically, we introduce a semantic correspondence convolution (SCC) module to dynamically compute the correspondences from deep and low resolution feature maps, and thus enhancing the discriminability of features. The experimental results demonstrate the effectiveness and the superior performance of our method on two benchmark datasets.

Source location with cross-coherence migration

Geophysics ◽  
2020 ◽  
Vol 85 (4) ◽  
pp. KS127-KS138 ◽  
Author(s):  
Yujin Liu ◽  
Yue Ma ◽  
Yi Luo
Keyword(s):  
Field Data ◽  
Random Noise ◽  
Synthetic Data ◽  
Seismic Energy ◽  
Seismic Source ◽  
Spectral Amplitude ◽  
Multiple Sources ◽  
Passive Seismic ◽  
Similar Images ◽  
Resolution Property

Locating microseismic source positions using seismic energy emitted from hydraulic fracturing is essential for choosing optimal fracking parameters and maximizing the fracturing effects in hydrocarbon exploitation. Interferometric crosscorrelation migration (ICCM) and zero-lag autocorrelation of time-reversal imaging (ATRI) are two important passive seismic source locating approaches that are proposed independently and seem to be substantially different. We have proven that these two methods are theoretically identical and produce very similar images. Moreover, we have developed cross-coherence that uses normalization by the spectral amplitude of each of the traces, rather than crosscorrelation or deconvolution, to improve the ICCM and ATRI methods. The adopted method enhances the spatial resolution of the source images and is particularly effective in the presence of highly variable and strong additive random noise. Synthetic and field data tests verify the equivalence of the conventional ICCM and ATRI and the equivalence of their improved versions. Compared with crosscorrelation- and deconvolution-based source locating methods, our approach shows a high-resolution property and antinoise capability in numerical tests using synthetic data with single and multiple sources, as well as field data.

scholarly journals Intelligent Ship Detection in Remote Sensing Images Based on Multi-Layer Convolutional Feature Fusion

2020 ◽  
Vol 12 (20) ◽  
pp. 3316 ◽  
Author(s):  
Yulian Zhang ◽  
Lihong Guo ◽  
Zengfa Wang ◽  
Yang Yu ◽  
Xinwei Liu ◽  
...  
Keyword(s):  
Remote Sensing ◽  
Feature Fusion ◽  
Atmospheric Correction ◽  
Recall Rate ◽  
Google Earth ◽  
Superior Performance ◽  
Detection Accuracy ◽  
Remote Sensing Images ◽  
Feature Maps ◽  
Ship Detection

Intelligent detection and recognition of ships from high-resolution remote sensing images is an extraordinarily useful task in civil and military reconnaissance. It is difficult to detect ships with high precision because various disturbances are present in the sea such as clouds, mist, islands, coastlines, ripples, and so on. To solve this problem, we propose a novel ship detection network based on multi-layer convolutional feature fusion (CFF-SDN). Our ship detection network consists of three parts. Firstly, the convolutional feature extraction network is used to extract ship features of different levels. Residual connection is introduced so that the model can be designed very deeply, and it is easy to train and converge. Secondly, the proposed network fuses fine-grained features from shallow layers with semantic features from deep layers, which is beneficial for detecting ship targets with different sizes. At the same time, it is helpful to improve the localization accuracy and detection accuracy of small objects. Finally, multiple fused feature maps are used for classification and regression, which can adapt to ships of multiple scales. Since the CFF-SDN model uses a pruning strategy, the detection speed is greatly improved. In the experiment, we create a dataset for ship detection in remote sensing images (DSDR), including actual satellite images from Google Earth and aerial images from electro-optical pod. The DSDR dataset contains not only visible light images, but also infrared images. To improve the robustness to various sea scenes, images under different scales, perspectives and illumination are obtained through data augmentation or affine transformation methods. To reduce the influence of atmospheric absorption and scattering, a dark channel prior is adopted to solve atmospheric correction on the sea scenes. Moreover, soft non-maximum suppression (NMS) is introduced to increase the recall rate for densely arranged ships. In addition, better detection performance is observed in comparison with the existing models in terms of precision rate and recall rate. The experimental results show that the proposed detection model can achieve the superior performance of ship detection in optical remote sensing image.

Sign in / Sign up

Export Citation Format

Share Document