Method for automated stratigraphy interpretation from borehole images

Patent metadata

Field	Value
Publication number	US-11900658-B2
Application number	US-202017593011-A
Country	US
Kind code	B2
Filing date	Mar 11, 2020
Priority date	Mar 11, 2019
Publication date	Feb 13, 2024
Grant date	Feb 13, 2024

Abstract

Embodiments of the present disclosure are directed towards systems and methods for automated stratigraphy interpretation from borehole images. Embodiments may include constructing, using at least one processor, a training set of synthetic images corresponding to a borehole, wherein the training set includes one or more of synthetic images, real images, and modified images. Embodiments may further include automatically classifying, using the at least one processor, the training set into one or more individual sedimentary geometries using one or machine learning techniques. Embodiments may also include automatically classifying, using the at least one processor, the training set into one or more priors for depositional environments.

First claim

What is claimed is:

1. A method for automated stratigraphy interpretation from borehole images comprising:
constructing, using at least one processor, a training set of images corresponding to a borehole, wherein the training set includes one or more of synthetic images, real images, and modified images;
automatically classifying, using the at least one processor, the training set into one or more individual sedimentary geometries using a machine learning model that has been trained based on images generated from wells with multiple deviations to automatically recognize one or more sedimentary geometries from one or more borehole images, regardless of borehole deviation, wherein the automatically classifying comprises:
identifying a longer than standard borehole image in the training set of images; and
applying a sliding window as a spatial sampling technique based on the identifying the longer than standard borehole image, wherein the spatial sampling technique includes providing a plurality of cropped images, corresponding to the sliding window, from the longer than standard borehole image as inputs to the machine learning model; and

automatically classifying, using the at least one processor, the training set into one or more priors for depositional environments, wherein the automatically classifying into the one or more priors includes:
building one or more tables of sedimentary geometry successions that represent one or more depositional environments; and
automatically obtaining, using the one or more tables, depositional environments from the training set of images.

2. The method of claim 1 , wherein the automatically classifying into the one or more priors for the depositional environments includes applying one or more machine learning techniques.

3. The method of claim 1 , wherein an addition of noise includes at least one of adding one or more masking stripes on one or more synthetic images of the synthetic images, adding one stripe on the one or more synthetic images, adding a one-pixel stripe to the one or more synthetic images, adding white noise to the one or more synthetic images, translating patterns on the one or more synthetic images, truncating the one or more synthetic images, or adding geometric noise.

4. The method of claim 1 , further comprising:
utilizing one or more automated individual sedimentary geometry predictions to establish a depositional environment predictor.

5. The method of claim 4 , wherein the depositional environment predictor includes a decision tree-based machine-learning, fuzzy-logic based algorithms, or a probabilistic graphical model.

6. A system for automated stratigraphy interpretation from borehole images comprising:
a memory configured to store one or more borehole images;
at least one processor configured to:
construct a training set of images corresponding to a borehole, wherein the training set includes one or more of synthetic images, real images, and modified images;
automatically classify the training set into one or more individual sedimentary geometries using a machine learning model that has been trained based on images generated from wells with multiple deviations to automatically recognize one or more sedimentary geometries from one or more borehole images, regardless of borehole deviation, wherein the automatically classifying comprises:
identifying a longer than standard borehole image in the training set of images; and
applying a sliding window as a spatial sampling technique based on the identifying the longer than standard borehole image, wherein the spatial sampling technique includes providing a plurality of cropped images, corresponding to the sliding window, from the longer than standard borehole image as inputs to the machine learning model;

automatically classify the training set into one or more priors for depositional environments, wherein the automatically classifying into the one or more priors includes:
building one or more tables of sedimentary geometry successions that represent one or more depositional environments; and
automatically obtaining, using the one or more tables, depositional environments from the training set of images.

7. The system of claim 6 , wherein constructing the training set includes a forward model to generate the synthetic images.

8. The method according to claim 1 , wherein constructing the training set includes a forward model to generate the synthetic images.

9. The method according to claim 8 wherein constructing the training set further includes an addition of noise to the synthetic images.

10. The system of claim 7 , wherein constructing the training set further includes an addition of noise to the synthetic images.

11. The system of claim 6 , wherein the automatically classifying into the one or more priors for the depositional environments includes applying one or more machine learning techniques.

12. The system of claim 6 , wherein an addition of noise includes at least one of adding one or more masking stripes on one or more synthetic images of the synthetic images, adding one stripe on the one or more synthetic images, adding a one-pixel stripe to the one or more synthetic images, adding white noise to the one or more synthetic images, translating patterns on the one or more synthetic images, truncating the one or more synthetic images, or adding geometric noise.

13. The system of claim 6 , further comprising:
utilizing one or more automated individual sedimentary geometry predictions to establish a depositional environment predictor.

14. The system of claim 13 , wherein the depositional environment predictor includes a decision tree-based machine-learning, fuzzy-logic based algorithms, or a probabilistic graphical model.

Assignees

• Schlumberger Technology Corp

Inventors

• Lefranc Marie
• Bayraktar Zikri
• Kristensen Morten
• Marza Philippe
• Le Nir Isabelle
• Prange Michael
• Kherroubi Josselin

Classifications

• G06V10/772 Primary

  Determining representative reference patterns, e.g. averaging or distorting patterns; Generating dictionaries · CPC title

• G06F18/2148

  characterised by the process organisation or structure, e.g. boosting cascade · CPC title

• G06F18/2415

  based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate · CPC title

• G06F18/2431

  Multiple classes · CPC title

• G06V10/774

  Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting · CPC title