Image-based analysis of a geological thin section

Patent metadata

Field	Value
Publication number	US-9626771-B2
Application number	US-201615082540-A
Country	US
Kind code	B2
Filing date	Mar 28, 2016
Priority date	May 20, 2015
Publication date	Apr 18, 2017
Grant date	Apr 18, 2017

Abstract

Techniques for an image-based analysis of a geological thin section include (i) acquiring a plurality of images from a geological thin section of a rock sample from a subterranean zone; (ii) manipulating the plurality of images to derive a composite image; (iii) optimizing the composite image to derive a seed image; (iv) identifying, in the seed image, a particular seed pixel of a plurality of contiguous pixels that comprise an image of a grain of a plurality of grains of the rock sample in the seed image; (v) determining, with a specified algorithm, a shape of the grain based on the seed pixel; (vi) determining, based on the shape of the grain, a size of the grain; and (vii) preparing the determination of the size of the grain for presentation to a user.

First claim

What is claimed is:

1. A computer-implemented method for an image-based analysis of a geological thin section, comprising:
(i) acquiring a plurality of images from a geological thin section of a rock sample from a subterranean zone, wherein acquiring a plurality of images from a geological thin section of a rock sample from a subterranean zone comprises:
acquiring at least one plane-polarized image from the geological thin section by acquiring four plane-polarized images from the geological thin section, each plane-polarized image rotated at a distinct angle relative to a zero position angle, from the geological thin section, and
acquiring at least two cross-polarized images from the geological thin section by acquiring four cross-polarized images from the geological thin section, each cross-polarized image rotated at the distinct angle relative to the zero position angle, from the geological thin section;

(ii) manipulating the plurality of images to derive a composite image, wherein manipulating the plurality of images to derive the composite image comprises:
registering the plurality of images, the registering comprising determining a rotational center for each of the plurality of images about which to apply a rotational transformation,
applying the rotational transformation to each of the plurality of images with a rotation angle equal to an opposite value of the distinct angles relative to the zero position angle,
generating a first composite image,
generating a second composite image, and
generating a third composite image, wherein the first composite image is generated by applying an edge detection algorithm to the at least two cross-polarized images, and the first composite image comprises an aggregation of a plurality of first composite images, with each first composite image corresponding to a particular one of the plurality of cross-polarized images;

(iii) optimizing the composite image to derive a seed image;
(iv) identifying, in the seed image, a particular seed pixel of a plurality of contiguous pixels that comprise an image of a grain of a plurality of grains of the rock sample in the seed image;
(v) determining, with a specified algorithm, a shape of the grain based on the seed pixel;
(vi) determining, based on the shape of the grain, a size of the grain; and
(vii) preparing the determination of the size of the grain for presentation to a user.

2. The computer-implemented method of claim 1 , further comprising determining, based on the shape of the grain, at least one of a sphericity, a roundness, an elongation, or a sharpness of the grain.

3. The computer-implemented method of claim 1 , further comprising generating an updated composite image to remove the shape of the grain from the composite image by setting the plurality of contiguous pixels that comprise the image of the grain to nil values.

4. The computer-implemented method of claim 1 , wherein the distinct angles comprise 0 degrees, 25 degrees, 45 degrees, and 65 degrees from the zero position angle.

5. The computer-implemented method of claim 1 , wherein the second composite image comprises a combined cross-polarized image based on the first composite image, the combined cross-polarized image comprising an average cross-polarized image based on the plurality of first composite images.

6. The computer-implemented method of claim 5 , wherein the third composite image comprises a segmented image of the at least one plane-polarized image, the segmented image generated by:
determining a cut-off value of a color that distinguishes the grain of the geological thin section and a pore of the geological thin section, the determined cut-off value based on a color of an epoxy of the geological thin section; and
applying the cut-off value to the at least one plane-polarized image.

7. The computer-implemented method of claim 6 , wherein manipulating the plurality of images to derive a composite image comprises generating a final composite image used to derive the seed image by tagging grain edges of the plurality of grains in the third composite image based on the first composite image to create the final composite image.

8. The computer-implemented method of claim 7 , wherein optimizing the composite image to derive the seed image comprises:
generating a first seed image based on an average of the final composite image;
generating a second seed image based on an absolute deviation or a standard deviation of the first seed image; and
generating a third seed image by assigning, in the second seed image, an absolute deviation value for each pixel in the second seed image that comprises an average value lower than a threshold.

9. The computer-implemented method of claim 8 , further comprising selecting the seed pixel from the third seed image based on the lowest average value of the pixels.

10. The computer-implemented method of claim 9 , wherein the specified algorithm comprises a seeded region growing algorithm (SRG), the method further comprising:
adding the seed pixel to a sequential search list (SSL) queue; and
determining, with the SRG, a shape of the grain based on the seed pixel by:
selecting the seed pixel from the SSL queue; and
for each of a plurality of neighboring pixels adjacent the seed pixel:
determining a color similarity of the neighboring pixel to the seed pixel; and
based on the color similarity of the neighboring pixel meeting a threshold color similarity, adding the neighboring pixel to the grain shape.

11. The computer-implemented method of claim 10 , wherein determining a color similarity of the neighboring pixel to the seed image comprises measuring a Euclidean distance between a color vector of the seed pixel and a color vector of the neighboring pixel, and the threshold color similarity comprises a maximum value of the Euclidean distance.

12. The computer-implemented method of claim 1 , wherein determining, based on the shape of the grain, a size of the grain, comprises at least one of:
determining a length of a longest segment inside the grain; or
determining a diameter of a circumscribed circle of the grain.

13. The computer-implemented method of claim 1 , further comprising:
executing an iterative process by repeating steps (iii)-(vi) for another grain of the plurality of grains of the rock sample; and
stopping the iterative process when the determined grain size of a particular grain of the plurality of grains is less than a specified threshold grain size.

14. The computer-implemented method of claim 1 , wherein the rock sample comprises an anisotropic rock sample or a clastic rock sample.

15. A system for an image-based analysis of a geological thin section, comprising:
a polarizing microscope; and
a control system that comprises a memory and one or more processors, the memory comprising instructions operable when executed by the one or more processors to perform operations comprising:
(i) acquiring a plurality of images from a geological thin section of a rock sample from a subterranean zone, wherein acquiring a plurality of images from a geological thin section of a rock sample from a subterranean zone comprises:
acquiring at least one plane-polarized image from the geological thin section by acquiring four plane-polarized images from the geological thin section, each plane-polarized image rotated at a distinct angle relative to a zero position angle, from the geological thin section, and
acquiring at least two cross-polarized images from the geological thin section by acquiring four cross-polarized images from the geological thin section, each cross-polarized image rotated at the di

Assignees

• Saudi Arabian Oil Co

Inventors

• Mezghani Mokhles Mustapha
• Shammari Salem Hamoud

Classifications

• G06T7/11 Primary

  Region-based segmentation · CPC title

• G06T7/60 Primary

  Analysis of geometric attributes · CPC title

• H04N23/631

  Graphical user interfaces [GUI] specially adapted for controlling image capture or setting capture parameters · CPC title

• G06F18/24

  Classification techniques · CPC title

• G06T3/60

  Rotation of whole images or parts thereof · CPC title