Hopfield neural network for seismic horizon picking

Huang, Kou Yuan

doi:10.1190/1.1885963

Cited by 12 publications

(4 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For automatic horizon tracking, various horizon trackers have been developed and implemented (e.g., Harrigan et al . ; Leggett, Sandham and Durrani ; Huang ; Faraklioti and Petrou ; Patel et al . ; Yu et al .…”

Section: Application Example: Horizon Modellingmentioning

confidence: 99%

“…Performing horizon tracking on the sorted seeds leads to a volume of all horizons. For automatic horizon tracking, various horizon trackers have been developed and implemented (e.g., Harrigan et al 1992;Leggett, Sandham and Durrani 1996;Huang 1997;Faraklioti and Petrou 2004;Patel et al 2010;Yu et al 2011;Di et al 2018). This study uses the dip vector-guided method (Di et al 2018) to grow the tracking from the given seed through the entire survey, and the dip vector is prepared using the proposed waveform curvature analysis (Fig.…”

Section: Application Example: Horizon Modellingmentioning

confidence: 99%

See 1 more Smart Citation

Three‐dimensional curvature analysis of seismic waveforms and its interpretational implications

Alfarraj

AlRegib

2018

Geophysical Prospecting

View full text Add to dashboard Cite

A B S T R A C TThe idea of curvature analysis has been widely used in subsurface structure interpretation from three-dimensional seismic data (e.g., fault/fracture detection and geomorphology delineation) by measuring the lateral changes in the geometry of seismic events. However, such geometric curvature utilizes only the kinematic information (two-way traveltime) of the available seismic signals. While analysing the dynamic information (waveform), the traditional approaches (e.g., complex trace analysis) are often trace-wise and thereby fail to take into account the seismic reflector continuity and deviate from the true direction of geologic deposition, especially for steeply dipping formations. This study proposes extending the three-dimensional curvature analysis to the waveforms in a seismic profile, here denoted as the waveform curvature, and investigates the associated implications for assisting seismic interpretation. Applications to the F3 seismic dataset over the Netherlands North Sea demonstrate the added values of the proposed waveform curvature analysis in four aspects. First, the capability of the curvature operator in differentiating convex and concave bending allows automatic decomposition of a seismic image by the reflector types (peaks, troughs and zero crossings), which can greatly facilitate computer-aided horizon interpretation and modelling from three-dimensional seismic data. Second, the signed minimum curvature offers a new analytical approach for estimating the fundamental and important reflector dip attribute by searching the orientation associated with least waveform variation. Third, the signed maximum curvature makes it possible to analyse the seismic signals along the normal direction of the reflection events. Finally, the curvature analysis promotes the frequency bands of the seismic signals and thereby enhances the apparent resolution on identifying and interpreting subtle seismic features.

show abstract

Section: Application Example: Horizon Modellingmentioning

confidence: 99%

Section: Application Example: Horizon Modellingmentioning

confidence: 99%

Three‐dimensional curvature analysis of seismic waveforms and its interpretational implications

Alfarraj

AlRegib

2018

Geophysical Prospecting

View full text Add to dashboard Cite

show abstract

“…Neural network based methods have a long history for these applications. Early works Harrigan et al [1992], Veezhinathan et al [1993], Liu et al [2005], Huang [2005], , Kusuma and Fish [2005], Alberts et al [2000] use multi-layer perceptron or recurrent networks of a few layers. Neither the quantity or quality of data, nor the computing power used in these early works were comparable with today's standards.…”

Section: Previous Work and Related Problemsmentioning

confidence: 99%

Multiresolution neural networks for tracking seismic horizons from few training images

2019

View full text Add to dashboard Cite

Detecting a specific horizon in seismic images is a valuable tool for geological interpretation. Because hand-picking the locations of the horizon is a time-consuming process, automated computational methods were developed starting three decades ago. Older techniques for such picking include interpolation of control points however, in recent years neural networks have been used for this task. Until now, most networks trained on small patches from larger images. This limits the networks ability to learn from large-scale geologic structures. Moreover, currently available networks and training strategies require label patches that have full and continuous annotations, which are also time-consuming to generate.We propose a projected loss-function for training convolutional networks with a multi-resolution structure, including variants of the U-net. Our networks learn from a small number of large seismic images without creating patches. The projected loss-function enables training on labels with just a few annotated pixels and has no issue with the other unknown label pixels. Training uses all data without reserving some for validation. Only the labels are split into training/testing. Contrary to other work on horizon tracking, we train the network to perform non-linear regression, and not classification. As such, we propose labels as the convolution of a Gaussian kernel and the known horizon locations that indicate uncertainty in the labels. The network output is the probability of the horizon location. We demonstrate the proposed computational ingredients on two different datasets, for horizon extrapolation and interpolation. We show that the predictions of our methodology are accurate even in areas far from known horizon locations because our learning strategy exploits all data in large seismic images.

show abstract

“…Horizon tracking using neural-networks has seen a few time-periods of varying activity [Harrigan et al,Figure 4: The predicted segmentation from Figure 3b (using network output-regularization) overlaid on the seismic input data. 1992, Veezhinathan et al, 1993, Liu et al, 2005, Huang, 2005, Kusuma and Fish, 2005, Alberts et al, 2005. Algorithms that are not based on learning have also made progress, see, e.g [Wu and Fomel, 2018] for recent work that combines and extends multiple concepts on deterministic horizon tracking.…”

Section: Horizon Tracking By Interpolation Of Scattered Picksmentioning

confidence: 99%

Neural networks for geophysicists and their application to seismic data interpretation

2019

View full text Add to dashboard Cite

Neural-networks have seen a surge of interest for the interpretation of seismic images during the last few years. Network-based learning methods can provide fast and accurate automatic interpretation, provided there are sufficiently many training labels. We provide an introduction to the field aimed at geophysicists that are familiar with the framework of forward modeling and inversion. We explain the similarities and differences between deep networks to other geophysical inverse problems and show their utility in solving problems such as lithology interpolation between wells, horizon tracking and segmentation of seismic images. The benefits of our approach are demonstrated on field data from the Sea of Ireland and the North Sea.1

show abstract

Hopfield neural network for seismic horizon picking

Cited by 12 publications

References 2 publications

Three‐dimensional curvature analysis of seismic waveforms and its interpretational implications

Three‐dimensional curvature analysis of seismic waveforms and its interpretational implications

Multiresolution neural networks for tracking seismic horizons from few training images

Neural networks for geophysicists and their application to seismic data interpretation

Contact Info

Product

Resources

About