Classifications

• • G—PHYSICS
  • G06—COMPUTING; CALCULATING OR COUNTING
  • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
  • G06V10/00—Arrangements for image or video recognition or understanding
  • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
  • G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
• • G—PHYSICS
  • G06—COMPUTING; CALCULATING OR COUNTING
  • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
  • G06V10/00—Arrangements for image or video recognition or understanding
  • G06V10/20—Image preprocessing
  • G06V10/255—Detecting or recognising potential candidate objects based on visual cues, e.g. shapes
• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
  • G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
  • G01N33/24—Earth materials
• • G—PHYSICS
  • G06—COMPUTING; CALCULATING OR COUNTING
  • G06F—ELECTRIC DIGITAL DATA PROCESSING
  • G06F18/00—Pattern recognition
  • G06F18/20—Analysing
  • G06F18/24—Classification techniques
  • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
  • G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
  • G06F18/24133—Distances to prototypes
• • G—PHYSICS
  • G06—COMPUTING; CALCULATING OR COUNTING
  • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
  • G06V10/00—Arrangements for image or video recognition or understanding
  • G06V10/20—Image preprocessing
  • G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
• • G—PHYSICS
  • G06—COMPUTING; CALCULATING OR COUNTING
  • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
  • G06V10/00—Arrangements for image or video recognition or understanding
  • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
  • G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
  • G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
• • G—PHYSICS
  • G06—COMPUTING; CALCULATING OR COUNTING
  • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
  • G06V10/00—Arrangements for image or video recognition or understanding
  • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
  • G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks

Definitions

• the present invention relates to the field of the detection of geological constituents of a rock sample.
• the characterization of a rock sample often requires the detection and counting of certain specific constituents present inside the sample, such as microfossils, nanofossils, plant debris, minerals, pollen grains, or analogues.
• certain specific constituents present inside the sample such as microfossils, nanofossils, plant debris, minerals, pollen grains, or analogues.
• this work can in particular be carried out by directly observing the rock sample with the naked eye, or by cutting a thin slice, called a "thin section", to observe it under an optical microscope. This makes it possible to categorize the rock more precisely and to estimate its physical properties, for applications in civil engineering, for the depollution of sites, for research into underground mineral or energy resources, etc.
• microfossils As for microfossils, one can consult, in particular, the Manuel de micropaléontologie published in 2011 by Mathieu, Bellier and Granier (ISBN 978-2-916733-04-3), as well as the database of foraminifera on the site Such a database includes a very large amount of data. This database can therefore only be used by an expert, and not in an automated process.
• microfossils For microfossils, one can refer in particular to: Beaufort, L, & Dollfus, D. (2004). Automatic recognition of coccoliths by dynamical neural networks. Marine Micropaleontology, 51 (1-2), 57-73. This method is also described in patent application WO2015132531 A1. However, such a method requires a large number of training images for the machine learning method. The patent application cites in particular the use of 10,000 images obtained by microscopy, and 100 images per morphological group. This large number of training images requires a very long preparation time and a very large number of preparatory operations: it is necessary to prepare the thin sections (in particular by cutting the rock, impregnation, and lapping) and carry out the observations medium of the microscope for each thin section.
• patent application EP 36737154 describes a system and method for determining the type of microfossils by applying machine learning.
• this method requires cutting the acquired images into a plurality of image data which requires user intervention.
• the aim of the present invention is to detect, classify and count geological constituents of a rock sample, automatically, with a limited preparation time and a limited number of operations (by taking a single image per sample of rock).
• the invention relates to a method for detecting and counting a geological constituent of a rock sample, by means of an automatic learning method of detection by location.
• the invention relates to a method for detecting and counting at least one geological constituent of a rock sample from learning images, each geological constituent belonging to a class of geological constituent.
• the following steps are implemented: a) Each geological constituent is surrounded on said learning images by means of a predefined geometric shape; b) Each predefined geometric shape on said learning images is associated with said class of said surrounded geological constituent; c) training an automatic detection-by-location learning algorithm to detect at least one geological constituent and to associate a class by means of said predefined geometric shapes and said associated classes of said learning images; d) An image of said rock sample is acquired; e) a localization detection model constructed from said automatic learning algorithm is applied to said acquired image of said sample of rock to surround at least one geological constituent by said predefined shape and to associate a class of geological constituent; and f) counting the number of geological constituents for each class for said acquired image of said rock sample.
• said classes of geological constituent are chosen from among microfossils, nanofossils, plant debris, minerals, pollen grains and the subdivisions of these elements.
• a color is associated for each class of geological constituent.
• said machine learning algorithm uses an artificial neural network, preferably a convolutional neural network, a fully convolutional neural network or a region-based convolutional neural network.
• said location-based detection model implements the following steps: i) generating at least one region of predefined geometric shape comprising a geological constituent and associating a class with each region of interest; ii) The position and/or at least one dimension of said region of interest is adjusted to surround said geological constituent.
• said location detection model generates at least one predefined geometric shape to surround a geological constituent.
• said predefined geometric shape is chosen from a square, a rectangle, an ellipse, a polygon, a circle.
• said image of said rock sample is acquired from a thin section of said rock sample.
• said image of said rock sample is acquired by means of an optical or electron microscope, with or without polarized light, a photograph, a tomography scanner with synchrotron, or a X-ray imaging.
• said method comprises at least one additional step chosen from:
• Portions of said acquired image of said rock sample comprising said at least one geological constituent are extracted and an image database is constructed of said at least one geological constituent with said portions extracted from said acquired image of said rock sample, or
• a supervised classification method is applied to categorize said at least one geological constituent
• At least one physical property of said rock is determined, or
• the invention also relates to a method for exploiting a soil or a subsoil, in which the following steps are implemented: a) At least one geological constituent of a rock sample is detected by means of the detection method according to one of the preceding characteristics; and b) said soil or said subsoil is exploited according to said detection of said geological constituent of said at least one rock.
• said exploitation of the ground or subsoil concerns the construction of a structure on said ground or in the subsoil, the storage of gas in the subsoil, or the exploitation of raw materials from said soil or said subsoil.
• -soil preferably said raw materials being the rock itself, or a material, or a fluid contained in said soil or in said subsoil.
• the invention relates to a method for determining the climate in a geographical area through the geological ages, in which the following steps are implemented: a) At least two rock samples are taken at different depths from a formation underground; b) At least one geological constituent is detected for each rock sample by means of the detection method according to one of the preceding characteristics; and c) said climate as well as the geological age in said geographical area are determined as a function of said at least one detected geological constituent.
• FIG. 1 illustrates the steps of the detection method according to one embodiment of the invention.
• Figure 2 illustrates a first variant of the predefined geometric shape used by the method according to the invention.
• Figure 3 illustrates a second variant of the predefined geometric shape used by the method according to the invention.
• FIG. 4 illustrates an image acquired from a rock sample as well as the geological constituents detected by means of the method according to one embodiment of the invention.
• the present invention relates to a method for detecting and counting geological constituents of a rock sample.
• a rock sample is a piece of rock, for example (non-limiting) taken by coring.
• geological constituents is meant at least one constituent included within the rock.
• Such a geological constituent of rock can be chosen from the following classes of geological constituent: microfossils, nanofossils, plant debris, minerals, pollen grains, or any subdivision of these elements: that is to say different types of microfossils, different types of nanofossils, different types of plant debris, different types of minerals, different types of pollen grains, or any analogous element.
• the geological constituents can be of small dimensions, in particular of millimetric or even microscopic size.
• the detection of geological constituents corresponds to the identification of geological constituents within the rock sample.
• the method according to the invention also makes it possible to identify the class of geological constituents detected, and to count the number of geological constituents for each class.
• the detection method according to the invention is based on the use of an image of the rock sample and on the implementation of an automatic learning algorithm to automate this detection.
• the acquired image of the rock sample, and the set of learning images implemented for training the automatic learning algorithm can be of all types .
• the image may be that of an entire rock sample.
• the image may be that of a core taken from a subterranean formation.
• the image may be that of a thin section of the sample of rock.
• the samples observed can in particular be blades called "thin blades" of three types: blades of very thin materials of the order of 30 ⁇ m, thicker blades of the order of 100 ⁇ m, or polished sections whose thickness can go up to 10 mm. To be able to correctly observe these different types of samples under the microscope, it is important to have a flat surface and a very low roughness (almost smooth).
• the image can be obtained in different ways, for example by means of an optical or electronic microscope, with or without polarized light, by means of a scanner, by means of a photograph, by means of a tomography with synchrotron, or by means of X-ray imaging or any analogous system.
• the image can be obtained by a microscope from a thin section of the rock sample.
• a microscope from a thin section of the rock sample.
• the image can be obtained by a scanner of a core taken from an underground formation.
• the detection method comprises the following steps:
• the method according to the invention comprises the following steps:
• Step 1 and the associated sub-steps 1.1 to 1.3 can be done offline, once and beforehand.
• Steps 2 and 3, and the associated sub-steps 3.1 and 3.2, can be performed online, for each rock sample considered. In other words, if one wishes to detect the geological constituents of several rock samples (for example several thin sections of the same rock), one repeats steps 2 and 3 for each rock sample, the algorithm d machine learning that can be trained only once. The steps will be detailed later in the description.
• the steps of the method can be limited to:
• the method according to the invention can be implemented by means of a computer system, in particular a computer, in particular steps 1 .3, 2, and 3 (3.1 and 3.2).
• Figure 1 illustrates, schematically and in a non-limiting manner, the steps of the detection method according to one embodiment of the invention.
• the machine learning algorithm forms a detection model by localization, which has as input a rock sample image and as output the detection and classification of geological constituents.
• the method according to the invention may comprise a prior step of acquisition of the learning images, this step being able to be implemented in the same way as step 2 of acquisition of the image of the sample Rock.
• this preliminary step may consist of preparing thin sections and performing microscopy of these thin sections to form the learning images.
• the machine learning algorithm can be a supervised classification machine learning algorithm, for example an artificial neural network, preferably a convolutional neural network (CNN).
• CNN convolutional neural network
• FCNN Fully Convolutional Neural Network
• R-CNN Region-based Convolutional Neural Network
• Convolutional neural network, fully convolutional neural network, and region-based convolutional neural network are particularly suitable for image processing.
• the machine learning algorithm can be of any type, including a non-convolutional neural network, a random forest method, a support vector machine (SVM) method. , or a method of Gaussian processes.
• SVM support vector machine
• the machine learning algorithm can be an unsupervised classification algorithm.
• each geological constituent is surrounded on the training images by means of a predefined geometric shape.
• the predefined geometric shape delimits portions of the training images that include a geological constituent.
• the same predefined geometric shape can be used for all the geological constituents, and only the dimensions of the predefined geometric shape vary from one geometric shape to another according to the dimensions of the geological constituents.
• this step can be implemented by a user.
• the predefined geometric shape can be chosen from a square, a rectangle, an ellipse, a polygon, a circle, or any similar shape.
• the predefined geometric shape can be an ellipse. Indeed, such a shape makes it possible to surround the geological constituent as closely as possible, which facilitates the detection of the geological constituents.
• Figure 2 illustrates, schematically and in a non-limiting manner, two portions of a training image.
• the figure on the left is the portion of the training image.
• This portion 1 of the learning image comprises a geological constituent 2.
• the figure on the right is the portion of the learning image on which the predefined geometric shape 3 has been represented, here a rectangle, which surrounds the geological constituent 2.
• Figure 3 illustrates, schematically and in a non-limiting way, two portions of a learning image.
• the figure on the left is the portion of the training image.
• This portion 1 of the learning image comprises a geological constituent 2.
• the figure on the right is the portion of the learning image on which the predefined geometric shape 3 has been represented, here an ellipse, which surrounds the geological constituent 2.
• each predefined geometric shape of each learning image is associated with a class of geological constituent.
• the class of the geological constituent surrounded by the predefined geometric shape is identified (for example: microfossil, nanofossil, minerals, or the type of nanofossil, the type of minerals, the type of plant debris, etc.), and the class of the identified geological constituent is associated with the predefined geometric shape.
• the learning algorithm makes it possible to detect and classify the geological constituents.
• this step can be implemented by a user.
• a color can be associated with each class of geological constituent.
• the predefined geometric shape may be of the color corresponding to the class of geological constituent identified. In this way, the visualization of the different classes of geological constituent is facilitated.
• a parameter can be associated with each class of geological constituent.
• the parameter corresponding to the class of geological constituent can be associated with the predefined geometric shape.
• the automatic learning algorithm of detection by location is trained by means of the learning images according to the predefined geometric shapes and their associated geological constituent classes.
• each detected object is characterized by a geometric shape described by its coordinates and at least one descriptor, such as in particular at least one dimension, as well as by a class.
• the training of the learning algorithm makes it possible to build a detection model by location of the geological constituents.
• the training of the automatic learning algorithm can be implemented in particular by one of the following methods:
• the method consists of applying a neural network to an entire image. This can divide the image into a grid of 19x19 regions and can predict five boxes locating the objects of interest per cell of the grid. A classification is also predicted for each of the boxes. A total of 1805 boxes are predicted per image but many do not contain an object or are redundant between them. A filter is finally applied to all the predictions to keep only the relevant boxes. These boxes represent the objects of interest in the image.
• This "one-step” method improves the inference performance of detection.
• This step involves acquiring an image of the rock sample to be analyzed.
• image acquisition methods adapted to the method according to the invention: optical or electronic microscope, polarized or not, photography, scanner, tomography with synchrotron, X-ray imaging, etc.
• this step may include a sub-step of preparing a thin section from the sample of rock to be analyzed.
• the geological constituents of the rock sample are automatically detected, classified and counted, by means of the acquired image of the rock sample, and by means of a detection model by location constructed from machine learning algorithm.
• a localization detection model formed from the machine learning algorithm is applied to the acquired image of the rock sample.
• a single acquired image of the rock sample is sufficient to detect the geological constituents.
• this step and the previous one can be implemented for several thin sections of the same rock, in order to improve the representativeness of the detection. and counting the geological constituents of the rock.
• the localization detection model formed by the automatic learning algorithm obtained in step 1.3 is applied to the image acquired of the rock sample in step 2.
• the predefined geometric shape is identical to the predefined geometric shape used in step 1.1; only the dimensions of the predefined geometric shape vary from one geological constituent to another to adapt to the dimensions of the geological constituent.
• a new image can be formed which includes the acquired image of the rock sample, on which the predefined geometric shapes are superimposed.
• the color or the parameter can also be assigned to each predefined geometric shape identified on the acquired image of the rock sample.
• the visualization and counting of the different classes of geological constituent of the acquired image of the rock sample is facilitated.
• the number of geological constituents is automatically counted for each class of geological constituent for the acquired image of the rock sample.
• Figure 4 illustrates, for a non-limiting illustrative example, in its left part, an image 4 acquired from a rock sample, and in its right part the image 9 of image 4 with the predefined geometric shapes surrounding the constituents geological.
• This acquired image 4 was obtained by microscopy of a thin section of the rock sample.
• Image 9 was obtained by the detection and counting method according to the invention.
• the machine learning algorithm was trained with 15 training images.
• a convolutional neural network and the “two-step” method were implemented.
• the geological constituents are identified by rectangles (predefined geometric shape) 5, 6, 7 and 8. Each type of rectangle corresponds to a class of foraminifer (which is a type of microfossils).
• the method according to the invention makes it possible to determine the following distribution:
• the method according to the invention makes it possible to automatically detect and count the geological constituents of an image of a rock sample.
• the method according to the invention may comprise at least one of the following additional steps:
• the proportion of the volume of the rock sample occupied by the geological constituents is determined from the detection of the geological constituent, in other words from the acquired image, and/or
• the morphological and/or textural characteristics (for example the shape, the dimensions, the presence of streaks, etc.) of the geological constituents are estimated, for example by means of image processing, and/or
• Portions of the acquired image of the rock sample comprising a geological constituent are extracted, and a database of images of the geological constituents is constructed from the portions extracted from the acquired image of the rock sample , and or
• a supervised classification method is applied to more finely categorize the geological constituents from the detection of the geological constituent. For example, one can determine the subspecies to which each detected microfossil belongs. This step may require having previously trained an artificial neural network to carry out this categorization, from a dedicated image database, and independently of the learning algorithm of the detection and counting method according to the invention, and/or
• At least one physical property of the rock is determined, in fact, the detection and counting of the geological constituents makes it possible to determine the physical properties in particular the porosity, the compaction, the mechanical resistance, or the permeability according to the type and the shape of detected geological constituents, and/or
• a deformation of the geological constituent is determined, which took place between the moment of the deposit of the geological constituent and the moment of the analysis of the rock sample. For example, one can analyze the shape of the geological constituent and compare its shape with other geological constituents belonging to the same class. These steps can be carried out automatically, preferably using computer means, such as a computer.
• the invention relates to a method for exploiting the ground or subsoil.
• the following steps are implemented: a) detecting at least one geological constituent of a rock sample by means of the detection method according to any one of the variants or combinations of variants described previously; and b) The soil and/or subsoil is mined according to the geological constituents of the rock sample detected in the previous step.
• the first step makes it possible in particular to determine the physical properties of the rock to be analyzed, the exploitation being implemented according to these physical properties.
• the method may comprise a preliminary step of taking a sample of rock from the soil or subsoil.
• the operation may relate in particular to the field of the construction of buildings or engineering structures, or the field of the exploitation of raw materials, or the field of gas storage, or the field of risk determination, or the field of site depollution, etc.
• the constitution of the rocky outcrops and/or the subsoil is determined by the categorization of the rock, and the construction is carried out by adapting in particular the foundations, the structure of the construction according to the categorization of the rock.
• the rock to be categorized can be taken from the ground or from the subsoil in a shallow manner.
• the constitution of rocky outcrops and/or the subsoil is determined by the detection of the geological constituents of the rock, and the exploitation of raw materials is carried out (the raw materials can be the rock itself, a material, for example a metal, or a fluid, for example hydrocarbons, present in the subsoil), by allowing including determining the appropriate areas (i.e. drilling areas, areas to be excavated for mines or quarries, etc. with the aim of recovering raw materials), determining the methods and tools to use (e.g. enhanced oil recovery, drilling tools, nature of explosive devices for mines or quarries, etc.).
• the rock to categorize can be taken from deep underground, can result from drill cuttings, or can come from a rock outcrop, etc.
• the constitution of the subsoil is determined by the categorization of the rock, and the gas storage is carried out in the subsoil in an appropriate zone, i.e. i.e. in an underground area capable of storing the gas without leakage.
• the constitution of a rocky outcrop (cliff) is determined by the categorization of the rock, and a consolidation operation is carried out if there is a risk of subsidence or landslide of the rock. outcrop.
• this process makes it possible to implement the exploitation of the soil and/or the subsoil, in a simple and rapid manner, without calling on an expert geologist.
• the process also makes it possible to process very large quantities of rock more quickly.
• the invention also relates to a method for determining the climate in a geographical area through the geological ages, in which the following steps are implemented: a) At least two rock samples are taken at different depths from an underground formation, rock samples that may come from the same succession of rock deposits; b) At least one geological constituent is detected for each rock sample by means of the detection method according to any one of the variants or combinations of variants described previously; and c) determining said climate as well as the geological age in the geographical area as a function of said at least one detected geological constituent.

Landscapes

• Engineering & Computer Science (AREA)
• Theoretical Computer Science (AREA)
• Physics & Mathematics (AREA)
• General Physics & Mathematics (AREA)
• Health & Medical Sciences (AREA)
• Life Sciences & Earth Sciences (AREA)
• Multimedia (AREA)
• General Health & Medical Sciences (AREA)
• Artificial Intelligence (AREA)
• Computer Vision & Pattern Recognition (AREA)
• Evolutionary Computation (AREA)
• Computing Systems (AREA)
• Software Systems (AREA)
• Medical Informatics (AREA)
• Databases & Information Systems (AREA)
• Chemical & Material Sciences (AREA)
• Data Mining & Analysis (AREA)
• General Life Sciences & Earth Sciences (AREA)
• Immunology (AREA)
• Geology (AREA)
• Food Science & Technology (AREA)
• Medicinal Chemistry (AREA)
• Analytical Chemistry (AREA)
• Biochemistry (AREA)
• Remote Sensing (AREA)
• Pathology (AREA)
• Environmental & Geological Engineering (AREA)
• Bioinformatics & Cheminformatics (AREA)
• Bioinformatics & Computational Biology (AREA)
• Evolutionary Biology (AREA)
• General Engineering & Computer Science (AREA)
• Image Analysis (AREA)
• Analysing Materials By The Use Of Radiation (AREA)

Applications Claiming Priority (2)

Publications (1)

Family

ID=74125314

Family Applications (1)

Country Status (5)

Family Cites Families (1)

• 2020
  • 2020-09-02 FR FR2008909A patent/FR3113735A1/fr active Pending
• 2021
  • 2021-08-16 US US18/042,660 patent/US20230316713A1/en active Pending
  • 2021-08-16 EP EP21752574.0A patent/EP4208818A1/de active Pending
  • 2021-08-16 WO PCT/EP2021/072698 patent/WO2022048891A1/fr unknown
  • 2021-08-16 CA CA3187737A patent/CA3187737A1/fr active Pending

Also Published As

Similar Documents

Legal Events