Wenchao Chen

On analysis-based two-step interpolation methods for randomly sampled seismic data

Interpolating the missing traces of regularly or irregularly sampled seismic record is an exceedingly important issue in the geophysical community. Many modern acquisition and reconstruction methods are designed to exploit the transform domain sparsity of the few randomly recorded but informative seismic data using thresholding techniques. In this paper, to regularize randomly sampled seismic data, we introduce two accelerated, analysis-based two-step interpolation algorithms, the analysis-based FISTA (fast iterative shrinkage-thresholding algorithm) and the FPOCS (fast projection onto convex sets) algorithm from the IST (iterative shrinkage-thresholding) algorithm and the POCS (projection onto convex sets) algorithm. A MATLAB package is developed for the implementation of these thresholding-related interpolation methods. Based on this package, we compare the reconstruction performance of these algorithms, using synthetic and real seismic data. Combined with several thresholding strategies, the accelerated convergence of the proposed methods is also highlighted.

Monochromatic Noise Removal via Sparsity-Enabled Signal Decomposition Method

Monochromatic noise always interferes with the interpretation of the seismic signals and degrades the quality of subsurface images obtained by further processes. Conventional methods suffer from several problems in detecting the monochromatic noise automatically, preserving seismic signals, etc. In this letter, we present an algorithm that can remove all major monochromatic noises from the seismic traces in a relatively harmless way. Our separation model is set up upon the assumption that input seismic data are composed of useful seismic signals and single-frequency interferences. Based on their diverse morphologies, two waveform dictionaries are chosen to represent each component sparsely, and the separation process is promoted by the sparsity of both components in their corresponding representing dictionaries. Both synthetic and field-shot data are employed to illustrate the effectiveness of our method.

Noise attenuation of the geophysical data using the framelet transform

This paper is on the use of the framelet transformation for noise surpression of geophysical data. The main contribution lies in the application of the transformation for geophysical data. The results of the proposed approach are compared to the use of f-x deconvolution filtering and the curvelet transformation for the same purpose on seismic data. Visually, an improvement in noise reduction is detected.

A New Tiling Scheme for 2-D Continuous Wavelet Transform With Different Rotation Parameters at Different Scales Resulting in a Tighter Frame

A desirable property when dealing with frames is that the low and up bounds of frames should be close to each other. This can accelerate the convergence speed in computing the dual-frame and simplify the process of recovering the signal from its frame coefficients. In this letter, a new 2-D continuous wavelet transform tiling scheme, which uses different rotation factor sampling methods for different scales, is proposed. The upper/lower bound of the 2-D wavelet family based on this new tiling scheme is derived. The upper/lower bound of the 2-D Morlet wavelet family is estimated, and the results show that the 2-D wavelet family based on the new tiling scheme can provide a higher lower bound and a lower upper bound compared with the 2-D wavelet family based on the old tiling scheme.

GPR clutter noise separation by statistical independency promotion

In ground-penetrating radar (GPR) surveys, the GPR data are always contaminated by high-amplitude clutter noise which consists of the director transmitter to receiver signal (also called crosstalk) and antenna ringdown. The crosstalk and antenna ringdown overlay severely the shallow reflection signals. The commonly used approach to filter out the clutter noise is to subtract the mean trace which can give satisfactory results when imaging point reflectors. Unfortunately, the response from an interface may be diminished by using this approach. This paper proposes a novel technique to separate the clutter noise from GPR data which utilizes the statistical independency of clutter noise and reflection signals of plane objects. In this method, the GPR data are modeled as a linear combination of the clutter and reflection signals. With this multichannel signal assumption, a statistical approach based on independent component analysis (ICA) is given to separate the clutter noise and reflection signals. The capacity of the proposed method is illustrated using sand trunk test data and ground track filed data.

Abrupt Feature Extraction Via the Combination of Sparse Representations

The complexity of channels in 3D seismic data always makes detailed interpretation challenging. Similarly, definition of sand bars and beaches is also complicated as they are only partially visible when seismic amplitude is examined. To improve the imaging of these features, current interpretation workflows use advanced color and opacity based co-rendering techniques to merge multiple attributes information. Other than to enhance these sedimentary features in 3D views with combinations of different attributes, this paper proposes to separate their reflection waveforms directly from 3D imaging data by exploiting the waveform diversity mechanism. Our separation model is set up upon the assumption that seismic data are composed of coherent events and abrupt features (correspond to sedimentary features). According to their appearance in vertical sections, we model these two kinds of seismic features as linear structures and punctate structures respectively. Two appropriate waveform dictionaries are chosen, one of which is used for the representation of the coherent events and the other for the sedimentary features. The separation process is promoted by the sparsity of both waveform components in their corresponding representing dictionaries. The capacity of the proposed method is illustrated using modeling data and real 3D seismic data with complex depositional systems.

Comparisons of four methods used for seismic quality factors estimation

Summary We compare four Q-factor estimation methods, including logarithmic spectral ratio (LSR), centroid frequency shifting (CFS), peak frequency shifting (PFS), and wavelet envelope peak instantaneous frequency (WEPIF) methods. First of all, principals of these four methods for Q-factor estimation are described briefly. Then some performances are compared for them with wavelet independence, noise resistance, and resolution of thin beds. For different source wavelets, LSR method works well; WEPIF method has slight wavelet dependence; CFS and PFS methods have strong wavelet dependence. For random noise, Q-factors estimated by CFS and PFS methods show bigger errors and instability, and Q-factors estimated by LSR and WEPIF methods give small relative error and good stability for seismic data of high signal-to-noise ratio (SNR). The synthetic test of a wedge model indicates that CFS method produces the lowest resolution, LSR and PFS methods demonstrate a moderate high resolution, and WEPIF method provides the highest resolution. Taking aspects compared above into consideration, the WEPIF method is relatively better than other three methods in some extent.

Qualitative seismic attenuation parameter estimation based on prestack data in the continuous wavelet domain

AbstractBecause the seismic wave propagates through the subsurface, part of the elastic energy eventually ends up as heat energy. This phenomenon is known as absorption (or anelastic attenuation). The factors causing anelastic attenuation include fluid movement and grain boundary friction. The seismic quality factor (Q) quantifies the anelastic attenuation and is commonly used in assisting reservoir characterization. However, current Q-estimation approaches are mainly implemented on a poststack seismic volume. The Q-estimation approaches applied to poststack seismic data assume that the seismic data are normal incident reflections, and they do not consider the effect of the travel path on seismic attenuation. In theory, the attenuation degree of the low-frequency component should differ from the attenuation degree of the high-frequency component for large-offset seismic data. We have developed a method to qualitatively estimate seismic attenuation in the prestack seismic domain. A continuous wavelet trans...

Hybrid image denoising method based on non-subsampled contourlet transform and bandelet transform.

The second generation bandelet transform uses the two-dimensional (2D) separable wavelet transform to improve its image denoising and compression performance. However, the 2D separable wavelet transform is not a shift-invariant transform and therefore cannot capture geometric information well. The authors propose a hybrid image denoising method in which the 2D separable wavelet transform in the second generation bandelet transform is replaced with the non-subsampled contourlet transform. The results of the application of the proposed method to several greyscale and colour benchmark images contaminated with various levels of Gaussian white noise and Poisson noise indicate that the proposed method has good peak signal-to-noise ratio and visual quality performance.

A Fast 3rd Generation Coherence Algorithm Realizing Method

Summary The 3rd generation coherence algorithm, which was proved to be an excellent tool to detect faults and has good antinoise ability, was restricted by its large computational cost on constructing covariance matrix and eigen decomposition. We aim at this problem in this paper and propose an efficient realizing method which was based on recursion and power algorithm to accelerate the 3rd generation algorithm. The results on field data show that our method can greatly reduce calculation.