A U-Net Based Multi-Scale Deformable Convolution Network for Seismic Random Noise Suppression

The sparse representation of a seismic signal plays a significant role to suppress random noise in the seismic data. Dictionary learning (DL) is an effective tool to find a suitable sparse representation of a seismic signal. As an unsupervised DL, non-negative matrix factorization (NMF) has been widely used for attenuating random noise in seismic data during recent years. However, due to the nonconvexity of the NMF, most of the NMF methods might suffer from bad local minima, especially for seismic data with heavy noise. To solve this issue, we propose a workflow to suppress seismic random noise in the time-space domain, named self-paced nonnegative dictionary learning (SPNDL). The SPNDL is implemented by incorporating the self-paced learning (SPL) method with the NMF. Inspired by the learning way of humans and animals, the SPL selects the training samples automatically from easy to complex. Based on the SPL theory, a valid seismic data can be recovered by training the sample points with the high signal-to-noise ratio (SNR) values, and then gradually including the sample points with the low SNR values into NMF training. Our proposed SPNDL solves the bad local minima issue and obtains a robust denoised result. The potential of the proposed SPNDL method is demonstrated by both examples of a synthetic data and a field data.

Seismic data usually contain various types of noise. Noise comes from signal collection environment, equipment (current noise), micro seismic, etc. Noise reduces the signal- to-noise ratio and resolution of seismic records, and affects the accuracy of seismic processing and stratigraphic interpretation. Improving the signal-to-noise ratio of seismic data is the primary task of seismic data processing. In practical research, many algorithms for suppressing random noise interference are proposed. In order to achieve better seismic data denoising technology, this paper introduces a new algorithm: improved shearlet transform algorithm. The basic principle of this algorithm is to perform shearlet transform on noisy seismic data, calculate the L2 norm of shearlet coefficients in different directions at each scale according to the distribution law of effective signals at different scales, and remove random noise according to the adaptive threshold in shearlet domain varying with scale and direction. Note: This paper was accepted into the Technical Program but was not presented at IMAGE 2022 in Houston, Texas.

The suppression of random noise in seismic data is an essential step in the processing of seismic signals. However, as the exploration environment is becoming more and more complicated, the energy of valid signals is weaker and the signal to noise (SNR) of seismic data is much lower which bring great difficulty to seismic data processing and interpretation. In this paper, we propose a new seismic random noise suppression approach based on the non-subsampled contourlet transform (NSCT). Firstly, the noisy seismic data is decomposed into different sub-bands of frequency and orientation responses using NSCT. Then, we use an appropriate thresholding operator to denoise the sub-bands of NSCT coefficients. Finally, we perform the inverse NSCT transform on the denoised NSCT components to reconstruct the denoised seismic data. We use synthetic data examples to demonstrate the superior performance of the proposed approach over the wavelet transform based threshold denoising method.

Summary The random noise in seismic data seriously limits the extraction of information related to underground geological structure and lithology parameters. How to remove such noise sufficiently while keeping the fidelity of original seismic data is an important issue in the seismic data processing. In the poststack seismic profile, the most obvious feature is the structural feature and non-stationary feature. Inspired by it, we propose an improved non-local means (NLM) method with adaptive search window (ASWNLM). This method includes determining the direction of the search window by a correlation algorithm and the height of the search window by the instantaneous center frequency. Tests on synthetic data and real post-stack data prove our method can remove random noise as much as possible while preserving signal detail well even without a proper filtering parameter h.

Improving the signal-to-noise ratio and suppressing random noise in seismic data is critical for high-precision processing. Although deep learning-based algorithms have gained popularity as denoising methods, they suffer from poor generalization ability, resulting in high training set construction cost and computation cost. To address this problem, we propose an unsupervised learning-based denoising method that includes an improved denoising strategy based on local similarity and replacement, a corresponding training method, and an improved network based on UNet. Our training method takes advantage of network convergence and allows direct training on the test region, effectively solving the problems associated with denoising methods using generalization ability while improving training performance. In addition, our network is specifically designed for the training method and incorporates various improvements that could further enhance the training effectiveness. Our method outperforms traditional denoising methods, as demonstrated by tests on synthetic and field data, with superior performance in random noise attenuation and reflection event reconstruction.

Key technologies for complex surface seismic acquisition in the Sichuan Basin and their application effect

Digital images have progressed significantly in many areas but various types of noise still exist in real-world images such as Gaussian noise, Poisson noise, Salt-and-pepper noise so on. Although image denoising plays an important role to remove noise from images by using many techniques to preserve the important feature of an image but faces the problem of computational complexity, over smoothing. To address these limitations proposed paper introduces hybrid image denoising technique that combines the strengths of Wavelet transform and Non-Local Means (NLM) filtering, with enhanced Otsu thresholding. In proposed work, firstly the noisy image is denoised by wavelet-based Otsu thresholding and then NLM is employed to improve the enhance edge preservation of an image by eliminating the further noise. Kodak 24 dataset is used to test the images. Furthermore, we compare the proposed technique to the existing technique based on the performance metrics such as Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), and Root Mean Squared Error (RMSE). The Kodak24 dataset is used to evaluate the proposed hybrid denoising method. Experimental results show that the proposed technique outperforms existing approaches in terms of PSNR (34.86), SSIM (0.93) and RMSE (4.61) demonstrating its effectiveness in denoising and preserving structural image quality. Additionally, the FOM and VIF scores indicate the outstanding results of the hybrid strategy: both the FOM and VIF values were the best among all other assessed denoising methods, attained to 0.99 and 0.59, respectively.

This research presents a development of the affine projection algorithm (APA) in voice communication applications. A method of controlling the parameters of the APA is devised to improve the performance in cancelling various types of ambient noise that could possibly corrupt speech signals in voice communication systems. Indicators are used to identify the type of noise accompanying the target signal. Then the corrupted signal is processed in a noise a cancellation setup in such a way that three parameters of algorithm are changed according to the nature of the noise. The spreading of elements in the covariance matrix of the noise is used as an indicator for the type of noise so that the projection order, step-size and filter length are changed at the same time. This way the performance of the canceller is improved rendering lower estimation error with a moderate computational power. The method was tested under various types of noise and showed better convergence performance than the original APA as well as other commonly used algorithms in noise cancellation systems. The MSE of the proposed VPAPA method drops to -65 dB in steady state compared to -20 dB using NLMS and just below -30 dB using standard APA with projection order of 8, while the powerful RLS reaches around -60dB under the same environment. The method can be useful for clearer voice communication in variable environmental noise.

In seismic data processing, denoising is crucial for identifying the Earth’s subsurface structure. Erratic noise in seismic data is challenging because of its high amplitude and non-Gaussian distribution characteristics. Dictionary learning with KSVD requires a longer computational time due to its iterative nature at various stages. To address this, we propose a robust denoising DHHL method that utilises mini-batch dictionary learning, the quantum Harrow–Hassidim–Lloyd (HHL) algorithm, and quantum singular value transformation (QSVT) to attenuate random and erratic noise in seismic data. The HHL algorithm provides a significant estimate of the sparse coefficient. The HHL provides a theoretically efficient method for sparse coding with favorable asymptotic scaling, improving the potential scalability of the method. Mini-batch dictionary learning constructs hierarchical dictionaries, thereby enhancing the capture of signal features in seismic data. QSVT enables the direct transformation of singular values, leading to faster convergence and more accurate sparse representations for dictionary updating. Here, we use minimisation loss using ℓ<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2,1</sub> to provide group sparsity and preserve the signal features. The proposed DHHL method was applied to both synthetic and field datasets. The efficiency of the DHHL method was assessed using evaluation metrics such as the signal-to-noise ratio (SNR), mean squared error (MSE), and local similarity map. The quantitative and visual results indicate that the DHHL method achieved a higher SNR, minimised the MSE with less signal leakage, and reduced the computational time by accelerating sparse coding.

We have developed convolutional sparse coding (CSC) to attenuate noise in seismic data. CSC gives a data-driven set of basis functions whose coefficients form a sparse distribution. The noise attenuation method by CSC can be divided into the training and denoising phases. Seismic data with a relatively high signal-to-noise ratio are chosen for training to get the learned basis functions. Then, we use all (or a subset) of the basis functions to attenuate the random or coherent noise in the seismic data. Numerical experiments on synthetic data show that CSC can learn a set of shifted invariant filters, which can reduce the redundancy of learned filters in the traditional sparse-coding denoising method. CSC achieves good denoising performance when training with the noisy data and better performance when training on a similar but noiseless data set. The numerical results from the field data test indicate that CSC can effectively suppress seismic noise in complex field data. By excluding filters with coherent noise features, our method can further attenuate coherent noise and separate ground roll.

Nowadays, voice interaction is increasingly applied to smart home appliances. There are various types of noises in our real lives, which requires speech enhancement technology to deal with multiple noisy speech scenarios and to process them in real-time. Traditional technologies of speech noise reduction require estimating the noise power spectrum first, then estimating the spectrogram gain value of noise reduction, such as minima controlled recursive averaging (MCRA), which can only deal with stationary environmental noises but cannot estimate noises with serious fluctuations of the power spectrum within quite limited durations. A highly complicated deep-learning model can estimate the power spectrum of various types of noise, but it cannot meet the requirement of real-time processing due to the large number of parameters of these general models. In this paper, we proposed a method combining deep-learning technologies with traditional signal processing techniques to estimate the power spectrum of various types of noises by designing a new model with fewer parameters, tiny deep convolutional recurrent network (TDCRN), and computing the speech gain value with the power spectrum. The result of our experiment indicates that, compared with the traditional technology and complicated deep-learning model, the proposed method, with only 0.29M parameters, increases the PESQ by more than 0.6, the STOI by more than 0.2 and the wake-up rate by more than 6%.

The contribution of type and levels of community noises to the perception of loudness, noisiness, and annoyance is being explored in ongoing research projects. Particular attention has been given to the relationship between scaling method and the conclusions it is possible to draw from the data, especially with regard to observers' abilities to differentiate noises according to the above attributes. Procedures for calibrating and possibly standardizing the perceptual scales have been attempted with preliminary success. Various types of noises have been investigated: aircraft noise under different flying conditions, traffic noise, and a number of other community noises. Besides specific scaling problems, the investigations have addressed the specific problems of observers' abilities to differentiate among noises that are combined in a complex setting as well as their ability to integrate various noises into a unitary perception of loudness and annoyance. Since the experimental setups reflect real‐life situations, the answers to these latter questions are relevant to understanding the responses of the individual who is exposed to combinations of noises in his or her community.The contribution of type and levels of community noises to the perception of loudness, noisiness, and annoyance is being explored in ongoing research projects. Particular attention has been given to the relationship between scaling method and the conclusions it is possible to draw from the data, especially with regard to observers' abilities to differentiate noises according to the above attributes. Procedures for calibrating and possibly standardizing the perceptual scales have been attempted with preliminary success. Various types of noises have been investigated: aircraft noise under different flying conditions, traffic noise, and a number of other community noises. Besides specific scaling problems, the investigations have addressed the specific problems of observers' abilities to differentiate among noises that are combined in a complex setting as well as their ability to integrate various noises into a unitary perception of loudness and annoyance. Since the experimental setups reflect real‐life situ...

Medical images have become omnipresent in diagnosis and therapy. However, they can be affected by various types of noise that reduce image quality and make the final diagnostic decision difficult. The main objective of this research is to effectively remove the noise while preserving the important image characteristics. This paper proposes a novel approach for image denoising based on discrete wavelet transform (DWT) with the selection of the best decomposition level and mother wavelet. Then, the thresholding function is carried out in the detail coefficients. Optimal thresholding is done using new optimization techniques such as the crow search algorithm and social spider optimization techniques. Finally, the inverse of DWT is applied to reconstruct the denoised image. The proposed method is evaluated using peak signal to noise ratio, mean square error, and the structural similarity index measure. The experimental results show the efficiency of the optimization‐based denoising method over standard methods. Interesting results are obtained with all kinds of noise, and improvements about 30 dB can be reached with the Rician noise.

Megavoltage computed tomography (MVCT) is an essential imaging modality for verifying patient positioning in helical tomotherapy. However, its clinical application in daily anatomical monitoring and adaptive radiotherapy is hindered by inherent image artifacts and poor soft-tissue contrast. This issue is particularly pronounced in pelvic radiotherapy, where the intra- and interfraction anatomical variations necessitate high-quality image guidance to ensure precise dose delivery. We developed a deep-learning-based framework for MVCT enhancement to improve anatomicalvisualizationand facilitate accurate adaptive treatment planning for cervical cancer. This study analyzed a retrospective cohort of 170 patients with cervical cancer who underwent helical tomotherapy. The proposed deep-learning-based algorithm employed a generative adversarial network (GAN) that integrated deformable convolution and a self-attention mechanism (SADC-EGAN) to improve MVCT image quality. Comparative analyses were conducted against representative baseline methods, including U-Net, Attention U-Net, U-Net++, Swin-UNet, CycleGAN, and Pix2Pix. Model performance was assessed using quantitative metrics, including mean absolute error (MAE), peak signal-to-noise ratio (PSNR), structural similarity index measure (SSIM), and Fréchet inception distance (FID). The synthetic computed tomography (sCT) images generated by the proposed SADC-EGAN method demonstrated superior Hounsfield unit (HU) accuracy and structural similarity comparedto the original MVCT. Specifically, the MAE between the sCT and kilovoltage computed tomography (kVCT) was reduced to 36.32±6.69 HU, compared with 56.72±9.09 HU for MVCT. In terms of image quality, the sCT images exhibited notable enhancements over MVCT images, with higher PSNR (32.54±2.31 vs. 29.40±1.56dB), improved SSIM (0.93±0.01vs. 0.89±0.02), and substantially lower FID (66.68±22.31vs. 153.52±28.77). The proposed SADC-EGAN framework, integrating deformable convolutions and self-attention, effectively generated high-quality kVCT-like images from MVCT, improving both HU accuracy and image quality. This approachhas clinical potential to enable online adaptive helical tomotherapy for cervical cancer.

In recent years, seismic data processing based on deep convolutional neural networks (CNN) has made great progress. However, most of these methods rely on the feature information on the same scale and cannot make full use of the self-similarity of seismic data. In order to solve this problem, this paper proposes a novel Pyramid Attention Residual Neural Network (PARNet) for seismic data denoising. Specifically, the main framework of the network includes the residual block (ResBlock), the residual block with multi-core convolutional layer, the parallel space and channel attention (MSCARB) and the pyramid module(Pyramid Module). Among them, MSACRB can not only extract more abundant features, but also focus on the features of channel and spatial dimension, so as to achieve stronger feature representation. The pyramid module captures multi-scale features through dilated convolution with different expansion rates. At the same time, the global context module can capture the global information of the feature map. The combination of the above two modules can achieve the purpose of capturing multi-scale global context features. This method has been verified on synthetic seismic data and field seismic data. The experiments use PSNR and SSIM as evaluation indicators. A large number of experiments have demonstrated that PARNet has efficient denoising ability and a competitive advantage compared with the latest seismic data denoising methods.