Image synthesis with graph cuts: a fast model proposal mechanism in probabilistic inversion

  title={Image synthesis with graph cuts: a fast model proposal mechanism in probabilistic inversion},
  author={Thomas Zahner and T. Lochbuhler and Gr{\'e}goire Mari{\'e}thoz and Niklas Linde},
  journal={Geophysical Journal International},
Geophysical inversion should ideally produce geologically realistic subsurface models that explain the available data. Multiple-point statistics is a geostatistical approach to construct subsurface models that are consistent with site-specific data, but also display the same type of patterns as those found in a training image. The training image can be seen as a conceptual model of the subsurface and is used as a non-parametric model of spatial variability. Inversion based on multiple-point… 

Patch‐based iterative conditional geostatistical simulation using graph cuts

This paper proposes an iterative patch‐based algorithm which adapts a graph cuts methodology that is widely used in computer graphics, and shows that the proposed approach obtains significant speedups and increases variability between realizations.

Probabilistic inversion with graph cuts: Application to the Boise Hydrogeophysical Research Site

Several amendments to the original graph cut inversion algorithm are introduced and a first-ever field application by addressing porosity estimation at the Boise Hydrogeophysical Research Site, Boise, Idaho is presented.

Deep generative models in inversion: a review and development of a new approach based on a variational autoencoder

This contribution reviews the conceptual framework of inversion with DGMs and study the principal causes of the nonlinearity of the generative mapping, and identifies a conflict between two goals: the accuracy of the generated patterns and the feasibility of gradient-based inversion.

3 D Probabilistic Inversion of Oscillatory Hydraulic Tomography Data with Graph Cuts : a Synthetic Case

Inversion of hydro-geophysical data allows modelers to characterize subsurface properties with some uncertainty. On one hand, in groundwater applications, the resulting subsurface models are used to

Hydrogeological Model Selection Among Complex Spatial Priors

A full Bayesian methodology based on Markov chain Monte Carlo is proposed to enable model selection among 2-D conceptual models that are sampled using training images and concepts from multiple-point statistics to provide a consistent ranking of the competing conceptual models considered.

Geological modeling using a recursive convolutional neural networks approach

A new technique for multiple-point geostatistical simulation based on a recursive convolutional neural network approach (RCNN) is presented, which requires conditioning data and a training image that depicts the type of geological structures expected to be found in the deposit.

Training‐Image Based Geostatistical Inversion Using a Spatial Generative Adversarial Neural Network

This work introduces and evaluates a new training‐image based inversion approach for complex geologic media that relies on a deep neural network of the generative adversarial network (GAN) type to allow for efficient probabilistic inversion using state‐of‐the‐art Markov chain Monte Carlo (MCMC) methods.



Simulation of Earth textures by conditional image quilting

The original method developed in computer graphics has been modified to accommodate conditioning data and 3‐D problems, and the results, when compared with previous multiple‐point statistics (MPS) methods, indicate an improvement in CPU time by at least 50.

Summary statistics from training images as prior information in probabilistic inversion

SUMMARY A strategy is presented to incorporate prior information from conceptual geological models in probabilistic inversion of geophysical data. The conceptual geological models are represented by

Conditioning of Multiple-Point Statistics Facies Simulations to Tomographic Images

Geophysical tomography captures the spatial distribution of the underlying geophysical property at a relatively high resolution, but the tomographic images tend to be blurred representations of

Bayesian inverse problem and optimization with iterative spatial resampling

Measurements are often unable to uniquely characterize the subsurface at a desired modeling resolution. In particular, inverse problems involving the characterization of hydraulic properties are

The Necessity of a Multiple-Point Prior Model

The newly introduced multiple-point simulation (mps) algorithms borrow the high order statistics from a visually and statistically explicit model, a training image, and it is shown that mps can simulate realizations with high entropy character as well as traditional Gaussian-based algorithms, while offering the flexibility of considering alternative training images with various levels of low entropy structures.

The Direct Sampling method to perform multiple‐point geostatistical simulations

This work proposes to sample directly the training image for a given data event, making the database unnecessary, and shows its applicability in the presence of complex features, nonlinear relationships between variables, and with various cases of nonstationarity.

A Bayesian mixture‐modeling approach for flow‐conditioned multiple‐point statistical facies simulation from uncertain training images

Multiple‐point statistics (MPS) provides a systematic approach for pattern‐based simulation of complex discrete geologic objects from a conceptual training image (TI) as prior model. The TI contains

Multiple‐point geostatistics for modeling subsurface heterogeneity: A comprehensive review

A comprehensive review of multiple-point (MP) geostatistics, which includes sequentially simulating patterns instead of points and using different geological scenarios (training images) for dynamic data inversion.