Automated method and system for detecting segmented rock particles

Abstract

Systems and methods are provided for analyzing images of a sample, such as for rock particles obtained during drilling of a geological formation. The systems and techniques utilize a large foundation model (LFM) in the segmentation of rock particles. The LFM may receive an image (or image data) of rock particles as input and generate pixel-level image segmentation (i.e., each pixel of the image is classified) as a segmented image. Further, activity annotations may be provided in conjunction with a graphical user interface (GUI) to allow a user to interact with the image as well as selective segmentation of the image.

Description

[0001] Cross-reference paragraphs

[0002] This application claims the benefit of U.S. Provisional Application No. 63 / 514,671, filed on July 20, 2023, entitled “AUTOMATED METHOD AND SYSTEM TODETECT SEGMENT ROCK PARTICLES,” the disclosure of which is incorporated herein by reference. Technical Field

[0003] The present disclosure generally relates to methods and systems for analyzing sample images of rock particles, such as obtained during drilling of geological formations. In particular, the present disclosure relates to detecting instances of rock particles in sample images using automatic image processing. Background Art

[0004] During the drilling of an oil well or another effluent (particularly gas, steam or water), rock particles are brought to the surface after being cut from the geological formation by the drill bit and carried to the surface by the mud circulating in the wellbore. The rock particles can be analyzed to enable the creation of a detailed record of the geological formation of the wellbore (e.g., a master log). The detailed record can be a function of the wellbore depth and various wellbore information, such as the lithology of the geological formation, can be determined.

[0005] Typically, geologists analyze sample images to determine the properties of rock particles and, therefore, the lithology of the geological formation from which the rock particles were extracted. However, this work requires a great deal of time and is usually performed in a laboratory far from the drilling rig, which makes it less efficient to control the drilling process based on the analysis results. Furthermore, this work is highly subjective because it is based on human observation. Therefore, it is desirable to have an improved method for analyzing sample images.

[0006] This section is intended to introduce the reader to various aspects of the art that may be related to various aspects of the present disclosure, which are described and / or claimed below. This discussion is believed to be helpful in providing the reader with background information to facilitate a better understanding of the various aspects of the present disclosure. Therefore, it should be understood that these statements should be read in this light, and not as admissions of prior art. Summary of the invention

[0007] The following describes the summary of certain embodiments disclosed herein. It should be understood that these aspects are presented only to provide the reader with a brief overview of these specific embodiments, and these aspects are not intended to limit the scope of the present disclosure. In fact, the present disclosure may include multiple aspects that may not be described below.

[0008] Certain embodiments of the present disclosure include a system comprising one or more processors and a memory that is accessible by the one or more processors and stores instructions that, when executed by the one or more processors, cause the one or more processors to perform operations including receiving image data of an image of a rock sample from an imaging system, analyzing the image data via an artificial intelligence (AI) model trained using a plurality of images, and generating a segmented image based on the analysis of the image data by the AI ​​model, wherein the segmented image includes a representation of the rock sample of the image data.

[0009] Certain embodiments of the present disclosure include a computer-implemented method comprising receiving image data of an image of a rock sample from an imaging system, analyzing the image data via an artificial intelligence (AI) model trained using a plurality of images, and generating a segmented image based on the analysis of the image data by the AI ​​model, wherein the segmented image includes a representation of the rock sample of the image data.

[0010] Certain embodiments of the present disclosure include a device comprising a display, a processor communicatively coupled to the display, and a memory communicatively coupled to the processor, the memory storing instructions that, when executed, cause the processor to perform operations including generating a graphical user interface (GUI), generating an image of a rock sample corresponding to image data received from an imaging system for presentation on the display, receiving a first input via a first user interaction with the GUI, generating a visual icon for display on the display at a specific location on the image of the rock sample displayed on the display, wherein the specific location is determined based on the first input, receiving a second input via a second user interaction with the GUI, and in response to the second input, analyzing at least a portion of the image data via an artificial intelligence (AI) model trained using a plurality of images, and generating a segmented image based on analysis of at least a portion of the image data by the AI ​​model, wherein the segmented image includes a representation of at least a portion of the rock sample.

[0011] With respect to various aspects of the present disclosure, there may be various improvements of the above-mentioned features. More features may also be incorporated into these different aspects. These improvements and additional features may exist alone or in any combination. For example, the various features associated with one or more of the illustrated embodiments discussed below may be incorporated into any of the above-mentioned aspects of the present disclosure alone or in any combination. The brief overview given above is intended only to familiarize the reader with certain aspects and backgrounds of the embodiments of the present disclosure, and does not limit the subject matter claimed. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Various aspects of the present disclosure may be better understood upon reading the following detailed description and referring to the accompanying drawings, in which:

[0013] Figure 1 is a schematic diagram of a drilling apparatus including a geological analysis system according to aspects of the present disclosure;

[0014] Figure 2 According to aspects of the present disclosure, Figure 1 A flowchart of a method for generating detailed records by a system;

[0015] Figure 3 According to aspects of the present disclosure, Figure 1 An example of a first image analyzed by an analysis system;

[0016] Figure 4 According to aspects of the present disclosure, Figure 3 An example of a first segmented image generated from a first image;

[0017] Figure 5 According to aspects of the present disclosure, Figure 3 An example of a second segmented image generated from a first image;

[0018] Figure 6 According to aspects of the present disclosure, Figure 1 an example of a second image analyzed by the analysis system;

[0019] Figure 7 According to aspects of the present disclosure Figure 6 an example of a graphical user interface (GUI) for interacting with a second image;

[0020] Figure 8 According to aspects of the present disclosure, Figure 6 An example of a first segmented image generated from a second image;

[0021] Fig. 9 According to aspects of the present disclosure Figure 8 An example of a GUI for interacting with a first segmented image; and

[0022] Fig.10 According to aspects of the present disclosure, Figure 8 An example of generating a second segmented image from a first segmented image. DETAILED DESCRIPTION

[0023] One or more specific embodiments will be described below. In order to provide a concise description of these embodiments, not all features of the actual implementation are described in the specification. It should be understood that in the development of any such actual implementation, as in any engineering or design project, many implementation-specific decisions must be made to achieve the developer's specific goals, such as complying with system-related and enterprise-related constraints, which may vary from implementation to implementation. In addition, it should be understood that such development efforts may be complex and time-consuming, but for ordinary technicians who benefit from this disclosure, this is still a routine task of design, fabrication and manufacturing.

[0024] When introducing elements of various embodiments of the present disclosure, the articles "a," "an," and "the" are intended to indicate that there are one or more elements. The terms "comprising," "including," and "having" are intended to be inclusive, and mean that there may be additional elements in addition to the listed elements. Furthermore, it should be understood that reference to "one embodiment" or "an embodiment" of the present disclosure is not intended to be interpreted as excluding the existence of additional embodiments that also incorporate the recited features.

[0025] As used herein, the terms “connecting…,” “connecting,” “connected,” “connected with…,” and “connected…” are used to mean “directly connected with…” or “connected with…”; the term “group” is used to mean “one element” or “more than one elements.” In addition, the terms “connecting…,” “connecting…,” “coupled,” “coupled together,” and “connected with…” are used to mean “directly connected together” or “connected together via one or more elements.”

[0026] Furthermore, as used herein, the terms "real-time," "real-time," or "substantially real-time" may be used interchangeably and are intended to describe operations (e.g., computing operations) that are performed without any human-perceivable interruptions between operations. For example, data associated with the systems described herein may be collected, transmitted, and / or used in control calculations in "substantially real-time," such that data reading, data transmission, and / or data processing steps may occur once per second, once every 0.1 seconds, once every 0.01 seconds, or even more frequently during operation of the system (e.g., while the system is running). Furthermore, as used herein, the terms "automatic" and "automated" are intended to describe, for example, operations that are performed or caused to be performed solely by an analysis system without human intervention.

[0027] The present disclosure relates to a system and method for analyzing images of drill cuttings samples taken from geological formations by a drilling system. Drill cuttings samples are cut from geological formations during drilling and can be used to evaluate geological formations and characterize one or more characteristics thereof, such as mineralogy, lithology, porosity, density, etc., based on the location (e.g., X, Y, Z coordinates and / or depth) in the geological formation. For example, images of drill cuttings samples can be used to identify the lithology of rocks in the drill cuttings samples and predict characteristics and parameters of the geological formations. The lithology of the rock samples can be used to generate a detailed record (e.g., a master logging file) of the geological formation. The detailed record may include information about the geological characteristics (e.g., lithology, layer, depositional environment) and rock physical characteristics (e.g., water saturation, porosity, permeability, shale volume) of the geological formation, which can be used to control the drilling system or the drilling plan of the drilling system. However, as an initial step in using drill cuttings images to evaluate geological formations and characterize one or more characteristics thereof, the images should be segmented (i.e., binary classification of the images at the pixel level).

[0028] In one embodiment, identification of rock particles during drilling may be performed by an operator, for example, using a microscope or other tool. This may be a time-intensive process, and the results are often operator dependent (e.g., their level of experience, etc.). Thus, in another embodiment, identification of rock particles may be automated, for example, to reduce operating costs, to allow operators to focus on tasks that require more expertise.

[0029] A technique for automatically identifying rock particles includes the use of machine learning or deep learning. In this embodiment, a deep learning or neural network processor or computing system is trained, including a software-implemented neural network, a machine learning system, or a deep learning system, and one or more models are generated, and then one or more models are used to identify rock particles. The construction of this type of model may be a complex and / or lengthy process, and may include a large amount of annotation of rock particles by domain experts (with consistency checks, etc.), the design of the model, and subsequent testing. In addition, the trained model has difficulties in being generally applicable. That is, it may be difficult to generate a model that is universal for many types of strata (e.g., different geology) and / or robust to environmental changes (e.g., lighting, wet rock particles, etc.).

[0030] Therefore, according to another embodiment, alternative systems and techniques may be utilized whereby a large basis model (LFM) is utilized in the identification of rock particles. Such an LFM may generally be considered an artificial intelligence (AI) model trained on an image according to the present embodiment. That is, the present embodiment includes an LFM that receives an image of rock particles as input and generates a pixel-level image classification (i.e., each pixel of the image is classified). This allows for the identification of, for example, objects and / or the shape of objects. One advantage of this embodiment is that it can operate without any rock particle specific training (e.g., zero shot learning). However, in some embodiments, the model may be retrained using, for example, specific images of known rock particles. Furthermore, in some embodiments, active annotations may be provided as part of a graphical user interface (GUI) to allow a user to interact with the classified images.

[0031] Considering the above, Figure 1 An exemplary oil and gas work site 10 is shown having a geological analysis system 11 for analysis and control using a drilling system 12. Although an oil and gas work site 10 is described, it should be understood that the techniques described herein can be applied to other types of oil and gas production operations. The drilling system 12 includes a rotary drilling tool 14 and a surface device 18, such as for drilling a cavity 16, and a surface device 18, such as for placing drill pipes into the cavity 16. A wellbore 20 (e.g., a wellbore) defining the cavity 16 is formed in a bottom layer 21 by rotating the drilling tool 14. At the surface 22, a wellhead 23 having a discharge pipe 25 closes the wellbore 20. The drilling tool 14 includes a drilling head 27, a drill string 29, and a liquid injection head 31. The drilling head 27 includes a drill bit 33 for drilling through the rock of the bottom layer 21. The drilling head 27 is mounted at the lower portion of the drill string 29 and is located at the bottom of the wellbore 20. The drill string 29 includes a set of hollow drill pipes. These drill pipes define an inner space 35, which makes it possible to bring drilling fluid from the surface 22 to the drilling head 27. A liquid injection head 31 is mounted (e.g., threaded, bolted, etc.) to the upper part of the drill string 29. The drilling fluid includes drilling mud, such as water-based or oil-based drilling mud.

[0032] The surface device 18 includes a support 41 for supporting the drilling tool 14 and driving it to rotate, an injector 43 for injecting drilling fluid, and a shale shaker 45. The injector 43 is hydraulically connected to the injection head 31 to introduce and circulate the drilling fluid in the internal space 35 of the drill string 29. The shale shaker 45 collects the drilling fluid flowing out of the discharge pipe 25. The drilling fluid contains drilling residues, called rock particles (or cavings or cuttings). The shale shaker 45 includes a screen 46, which allows the separation of solid drilling rock particles, such as rock samples 47, from the drilling mud. The shale shaker 45 also includes an outlet 48 for discharging the rock sample 47. The rock sample 47 obtained at the outlet 48 has been cut from the geological formation during the drilling process and can be used to evaluate the geological formation and characterize one or more of its characteristics, such as its mineralogy, lithology, porosity, density, etc.

[0033] exist Figure 1 In the illustrated embodiment, the rock sample 47 may be automatically or manually sampled and transported to a conveyor 50, which may transport the rock sample 47 to a preparation device 52. The preparation device 52 may prepare the rock sample 47 before sending the rock sample 47 manually or automatically (e.g., via a conveyor) to an imaging system 54. The geological analysis system 11 may include all equipment associated with collecting, preparing, imaging, and analyzing the rock sample 47. For example, the geological analysis system 11 may include a shale shaker 45, a conveyor 50, a preparation device 52, an imaging system 54, and an analysis system 60. The preparation work may include washing, rinsing, drying, or sieving the rock sample, etc. The imaging system 54 may include an imaging device 56 to take an image of the rock sample 47. The imaging device 56 may be any type of optical or electronic microscope, camera, etc. The image obtained by the imaging device 56 may be a digital image, which may be automatically analyzed, as discussed in further detail below. The following examples are given with a camera that detects light in the visible spectrum, but the same methods can be applied to images captured with an infrared (IR) or ultraviolet (UV) camera that detects light in the UV or IR domain. The imaging system 54 can also include a control device 58 (e.g., a processor-based controller) to control the imaging device 56 and operating conditions associated with the image capture process within the imaging system 54 (e.g., lighting, temperature, humidity). For example, the control device 58 can adjust parameters of the imaging device 56 (e.g., focus, exposure, shutter speed, brightness and color, contrast, filters, resolution, zoom). The preparation device 52 and / or the imaging system 54 can be located at the oil and gas work site 10, or at one or more remote locations.

[0034] The analysis system 60 may be used to receive and analyze image data (e.g., digital images) from the imaging system 54, either directly or via a network 61. The analysis system 60 may be located at the oil and gas work site 10, or at one or more remote locations. The analysis system 60 may include a communication component 62, a processor 64, a memory 66, a data storage device 68, an input / output (I / O) port 70, a display 72, a prediction engine 74, and the like. The network 61 may include a transceiver, a receiver, and / or a transmitter to facilitate the transmission of data to and / or from the analysis system 60. For example, image data from the imaging system 54 may be transmitted to the analysis system 60 via the network 61. In addition, external data (e.g., data regarding geological formations) may be collected from a remote system and transmitted to the analysis system 60 via the network 61. However, in some embodiments, data may be transmitted directly from a device (e.g., the imaging system 54) to the analysis system 60. In fact, according to the present embodiment, the analysis system 60 may communicate with the device directly and / or via the network 61. In certain embodiments, data (e.g., image data) may be automatically transmitted from the imaging system 54 to the analysis system 60 for real-time analysis, thereby enabling real-time response (e.g., adjusting the imaging system 54, retaking an unacceptable image, adjusting the drilling system 12, etc.) to information obtained from the data analysis.

[0035] The communication component 62 can be a wireless or wired communication component (e.g., circuitry) that can facilitate communication between the analysis system 60, various types of devices, the network 61, etc. In addition, the communication component 62 can facilitate the transmission of data to the analysis system 60 so that the analysis system 60 can obtain the data from the analysis system 60. Figure 1 The communication component 62 may use various communication protocols, such as Open Database Connectivity (ODBC), TCP / IP, Distributed Relational Database Architecture (DRDA), Database Change Protocol (DCP), HTTP, other suitable current or future protocols, or combinations thereof.

[0036] The processor 64 may include a single-threaded processor, a multi-threaded processor, or both. The processor 64 may process instructions stored in the memory 66. The processor 64 may also include a hardware-based processor, each processor including one or more cores. The processor 64 may include a general-purpose processor, a special-purpose processor, or both. The processor 64 may be communicatively coupled to other internal components (e.g., the communication component 62, the data storage device 68, the I / O port 70, and the display 72).

[0037] The memory 66 and the data storage device 68 may be any suitable article of manufacture capable of serving as a medium for storing processor executable code, data, and the like. These articles of manufacture may represent computer readable media (e.g., any suitable form of memory or storage device) that may store processor executable code used by the processor 64 to perform the presently disclosed techniques. As used herein, an application may include any suitable computer software or program that may be installed on the analysis system 60 and executed by the processor 64. The memory 66 and the data storage device 68 may represent non-transitory computer readable media (e.g., any suitable form of memory or storage device) that may store processor executable code used by the processor 64 to perform the various techniques described herein. It should be noted that non-transitory simply means that the medium is tangible, rather than a signal.

[0038] The I / O port 70 may be an interface that may be coupled to other peripheral components such as input devices (e.g., keyboard, mouse), sensors, input / output (I / O) modules, and the like. The display 72 may operate as a human-machine interface (HMI) to depict visualizations associated with the software or executable code being processed by the processor 64. The display 72 may display a map of geological stratigraphic data corresponding to a location on a map (e.g., images and information derived from the images), warnings / alarms when image data is unacceptable, suggestions associated with warnings / alarms, and the like. In one embodiment, the display 72 may be a touch display capable of receiving input from an operator of the analysis system 60. The display 72 may be any suitable type of display, such as a liquid crystal display (LCD), a plasma display, or an organic light emitting diode (OLED) display. Furthermore, in one embodiment, the display 72 may be provided in conjunction with a touch-sensitive mechanism (e.g., a touch screen) that may be provided as part of a control interface for the analysis system 60.

[0039] It should be noted that the components described above with respect to the analysis system 60 are exemplary components, and the analysis system 60 may include more or fewer components as shown. In addition, although the components are described as part of the analysis system 60, the components may also be part of any suitable computing device described herein, such as the screen 46, the conveyor 50, the preparation device 52, the imaging device 56, the control device 58, and the analysis system 60, etc., to perform the various operations described herein.

[0040] In some embodiments, the prediction engine 74 can use various machine learning algorithms to analyze the images obtained for the rock sample 47 to identify the lithology of the rock sample. The prediction engine 74 can use one or more prediction models to analyze various data received by the analysis system 60. Various types of prediction models can be used to analyze data from various resources and generate prediction outputs. For example, the prediction engine 74 can be trained with supervised machine learning techniques, that is, the prediction model is trained with training data including input data and expected prediction output (e.g., labeled data sets). The prediction engine 74 can also be trained with unsupervised machine learning techniques, that is, the prediction model is trained with training data including input data but without expected prediction output (e.g., unlabeled data sets). The prediction engine 74 can include various types of artificial neural networks (ANNs), such as convolutional neural networks (CNNs), recurrent neural networks (RNNs), etc. The analysis system 60 can also communicate with a database 76, which can store information related to the oil and gas work site 10, the drilling system 12, related external resources (e.g., geological stratigraphic history), etc.

[0041] In other embodiments, the prediction engine 74 can be a large base model (LFM) for identifying rock particles, such as rock sample 47. Such an LFM can generally be considered an artificial intelligence (AI) model trained on images. In some embodiments, such training of the LFM does not require images of rock particles for training; alternative images can be used to train the LFM, and once trained, the LFM can be used to receive images of rock particles as input and generate pixel-level image classification (i.e., each pixel of the image is classified). This allows, for example, objects and / or shapes of objects to be identified, thereby providing segmentation of the input image (i.e., the rock particle image input to the LFM). In this way, the LFM can operate without any rock particle specific training (e.g., zero-shot learning). However, in some embodiments, the LFM can be retrained using, for example, specific images of known rock particles. The following is in conjunction with Figure 2 The use of the prediction engine 74 in conjunction with the LFM is described.

[0042] Figure 2 is used for Figure 1Flow chart of a computer-implemented method 78 for generating a detailed record (e.g., a master log) for a system. For example, the method 78 may be implemented using one or more processor-based systems (e.g., a processor-based controller) configured to control the drilling system 12, the imaging system 54, the analysis system 60, and / or related equipment at the oil and gas work site 10. At box 80, a drill cutting sample at a depth of the wellbore 20 may be received from the drilling system 12. At box 82, as described above, the shale shaker 45 may separate the solid drill cutting sample from the drilling mud by the screen 46 to obtain a rock sample 47 (i.e., rock particles). The rock sample 47 may be delivered (e.g., via the conveyor 50) to the preparation device 52, which may prepare the rock sample 47 before sending it to the imaging system 54. The preparation work may include flushing, rinsing, drying, or screening the rock sample, etc. At box 84, the imaging system 54 may capture an optical image of the rock sample 47 by controlling the imaging device 56 using the control device 58. At block 86, the analysis system 60 may receive an image of the rock sample from the imaging system 54 for analysis by, for example, a prediction engine 74 implemented using LFM. At block 88, the analysis system 60 may analyze the image of the rock sample. The analysis may include, for example, segmenting the image via the prediction engine 74, classifying individual rock grains from the rock sample 47 in the received image using the prediction engine 74 of the LFM to allow for subsequent use in the method 78.

[0043] At block 90, the analysis system 60 may use the analyzed image (e.g., the segmented image and / or individual rock particles selected therefrom) to identify the lithology of the rock sample. In some embodiments, as part of block 90, the user may also interpret the analyzed image and classify the rock particles into different lithologies.

[0044] After segmentation in box 88 and identification of the lithology of the rock sample 47 in box 90 (i.e., rock prediction), feature extraction in box 92 may be performed. Feature extraction in box 92 may include characterizing (i.e., converting) the segmented image information into a set of parameters (e.g., values) representing the color and texture information of each rock particle. Using the features (e.g., vectors), modeling, classification, clustering, etc. may be performed and the results may be used to control the drilling system 12 in box 94.

[0045] Figure 3 is an example of an image 96 to be analyzed by the analysis system 60 (more specifically, the prediction engine 74 utilizing LFM). As part of block 86, the image 96 may be generated and transmitted to the analysis system. Thereafter, in conjunction with block 88, the prediction engine 74 may perform segmentation on the image 96 utilizing LFM. The results of such segmentation are shown in FIG. Figure 4 shown.

[0046] Figure 4 An image 98 is shown, as from Figure 3 98 has been segmented via the prediction engine 74 using LFM. As shown, the image 98 includes rock particles that have been successfully segmented and rock particles that have not been successfully segmented. For example, drill cuttings 100, drill cuttings 102, drill cuttings 104, and drill cuttings 106 are examples of rock particles that have been successfully segmented in the image 98. Similarly, drill cuttings 108 and drill cuttings 110 are examples of rock particles that have not been successfully segmented in the image 98. That is, drill cuttings 108 and drill cuttings 110 are the same as their corresponding rock particles in the image 96.

[0047] One option to improve the accuracy of the segmented images generated by the prediction engine 74 using LFM is to adjust the level of refinement applied by the prediction engine 74. For example, Figure 4 The image 98 in can be generated with a default grid setting. This can be a default grid setting, such as 32×32, which generally corresponds to a coarse grid. The grid utilized can correspond to, for example, the number of initial points defined in the image (e.g., the area examined for drill cuttings features by the prediction engine 74). In addition, in some embodiments, an alternative grid setting with a higher resolution can be selected, such as 128×128 corresponding to a finer grid, or it can be set as the default grid setting. Using a finer grid, whether selected or set as default, can produce a segmented image with increased results (e.g., fewer rock particles in the image 98 that were not successfully segmented). Figure 5 An example of using a finer grid by the prediction engine 74 utilizing LFM is shown.

[0048] Figure 5 Shown as from Figure 3 An example of an image 112 of image 96 which has been segmented by prediction engine 74 using LEM has a higher resolution than that generated by Figure 4 1. The image 112 is a diagram of an image 112 that is a finer resolution than that applied by the prediction engine 74 of the LFM. As shown, the image 112 includes the drill cuttings 100, drill cuttings 102, drill cuttings 104, and drill cuttings 106 that have been successfully segmented into segmented rock particles. In addition, the drill cuttings 108 and drill cuttings 110 are also shown as having been successfully segmented into segmented rock particles. In this way, the image 112 illustrates the increased accuracy that can be obtained when a finer grid resolution is applied via the prediction engine 74. However, it should be noted that the image 112 is not as accurate as the image 112 generated by the prediction engine 74. Figure 4 The amount of time required to generate 98 images, using a finer grid resolution will also increase Figure 5 The amount of time required to create image 112.

[0049] In one embodiment, the prediction engine 74 using LFM may run a first segmentation at a first grid resolution (e.g., coarse). Thereafter, the user (or analysis system 60) may select a second (finer) grid resolution and re-run the segmentation on the image generated in the first segmentation. The level of grid resolution may be selected by the user, may be a default level, or may be selected by the analysis system 60, for example, based on a percentage or threshold of rock grains that were not characterized by the first segmentation. In other embodiments, the selection of the default and / or second grid resolution may be based on calculations, for example, a separate estimate of the grid size to be employed based on, for example, the number of elements in the image to be segmented or another factor. In another embodiment, a preliminary model (e.g., using a coarse grid resolution) may be applied as a first segmentation, and a second segmentation using a finer resolution grid may be applied to segment specific areas of the image (i.e., the user or analysis system 60 selects specific portions of the segmented image where the first segmentation was successful, and re-runs the segmentation on the remaining area with fewer representations). In this way, the processing time for the second segmentation may be accelerated because the area to be re-segmented is smaller than the original image.

[0050] Setting parameters for the model employed via the prediction engine 74 utilizing LFM can allow for variation (i.e., diversity) in the results and the time used to generate these results. An example of parameters that can be set includes how many initial points are defined in the image (i.e., the defined area of ​​the rock grains). In addition, if there is an initial estimate of the rock size corresponding to the rock grains, this can be applied to the grid defining the initial points. In addition, in some embodiments, if the fully automatic model (e.g., the base model) takes more time to process than expected, a light (light) specialized model (e.g., based on Mask R-CNN, as described in U.S. Patent Publication No. 2023 / 0220761 filed on January 7, 2022, the entire contents of which are incorporated herein by reference) can be implemented and customized for the task. The light model can be trained after the operation using high-quality labels (semi-) automatically generated from the base model (with a fine grid). Thereafter, the base model (with a fine grid resolution) can be used after the operation is completed, and time is less important. The results generated from the base model can provide high-quality labels (e.g., input) for retraining of the specialized model. In other embodiments, a base model (with a fine grid resolution) may be used in the operation, however, it may be used only on portions of the image where the dedicated model has high uncertainty (e.g., as a first segmentation when a coarse grid resolution is implemented as a first segmentation or as a selected second segmentation of less characterized areas).

[0051] In some embodiments, it may be advantageous to allow a user to interact with the image received from block 86 and / or the segmented image, for example, the segmented image generated using a prediction engine utilizing LFM as described above. Figure 6 An image 114 is shown as an example of an image that has been segmented in at least one of the above-described ways via the prediction engine 74 utilizing LFM. In some embodiments, it may be beneficial for a user to be able to interact with the image 114. Furthermore, it should be noted that reference is made to Figure 6-10 The described techniques and systems may alternatively be performed on images to be analyzed by the analysis system 60 (and more specifically, the prediction engine 74 utilizing the LFM) (e.g., images that have not yet undergone segmentation, such as Figure 3 Image 96).

[0052] Figure 6 The image 114 in includes drill cuttings 116. As shown, the drill cuttings 116 have been successfully segmented. As previously described, the analysis system 60 may include an I / O port 70, which can be connected to other peripheral components, such as input devices (e.g., keyboard, mouse), sensors, input / output (I / O) modules, etc. In addition, the display 72 can be operated as a human-machine interface (HMI) to depict visualizations associated with the software or executable code being processed by the processor 64. In one embodiment, the display 72 can be a touch display capable of receiving input from an operator of the analysis system 60. In this way, the display 72 can be provided in combination with a touch-sensitive mechanism (e.g., a touch screen), which can be used as part of a control interface for the analysis system 60.

[0053] Thus, the analysis system 60 may allow a user to interact, for example, launch, control, or operate a graphical user interface (GUI) or application running on the analysis system 60, and / or launch, control, or operate, for example, a portion of the platform executing on the analysis system 60. It will be appreciated that the GUI referred to above may be a user interface that allows a user to interact with the analysis system 60 in conjunction with the present platform through, for example, graphical icons, visual indicators, etc. Thus, the GUI may represent code or a program executed by the analysis system 60, for example, code or a program executed by the processor 64 interacting with one or more tangible, non-transitory machine-readable media (e.g., machine-readable media, such as memory 66) of the analysis system 60, which collectively store instructions, programs, etc. executable by the analysis system 60 to perform the methods and actions described herein, including providing a GUI and receiving and processing user input and / or other data input.

[0054] In some embodiments, the GUI allows for interaction with images displayed to the user (e.g., Figure 6The GUI can interact with the image 114 of the display. Thus, when the user makes a defined input (e.g., clicks a specific mouse button, a specific interaction with the display, presses one or more keys on a keyboard, etc.), the GUI can respond with a programmed response. Each response of the GUI can be programmed by the user to a specific input, and in some embodiments, the programming can be changed by the user (i.e., allowing user customization of the GUI). Figure 7 Examples of specific inputs and resulting GUI responses are shown.

[0055] Figure 7 An image 114 is shown. However, Figure 7 The image 114 of FIG. 1 is different in that the drill cuttings 116 have been removed from the segmented region due to a specific user input. For example, by inputting a left click on the mouse on a specific region of interest in the image 114, the segmented region (i.e., the drill cuttings 116) is removed. In addition, the GUI can add a positive point 118 to the user-selected region. In some embodiments, the positive point 118 can have a corresponding visual representation (e.g., a specific shape and / or color, such as yellow). The positive point 118 can represent, for example, a specific drill cutting (e.g., 116) to be segmented. Additional (second) user input (e.g., pressing the "Enter" key on the keyboard) can result in segmentation using the specific drill cutting (e.g., the drill cutting 116).

[0056] Figure 8 An image 120 is shown. Image 120 represents an image including segmentation using a particular drill cutting (e.g., drill cutting 116) after a second user input (e.g., pressing the "Enter" key on a keyboard). Other user inputs corresponding to actions taken by the prediction engine 74 using the LFM can be completed via the GUI. For example, in some embodiments, negative points can be introduced in place of positive points. Negative points can correspond to areas outside of a particular drill cutting. In some embodiments, negative points can be used if positive points are determined to be insufficient and there are additional rock particles to be incorporated. In addition, negative points can be automatically added using drill cutting neighbors (already segmented rocks can be used as negative guides).

[0057] Fig. 9An image 120 is shown in which a positive point 118 is added to an area selected by a user in conjunction with a user-defined user input (e.g., by inputting a left click on a mouse on a particular area of ​​interest in the image 120), and a segmented area (i.e., drill cuttings 122) is removed (if the area has already been segmented). In addition, the GUI can add the positive point 118 to the area selected by the user. Additional defined input by the user (e.g., inputting a right click on a mouse on a particular area of ​​interest in the image 120) will result in a negative point 124 being presented to the user. In some embodiments, the negative point 124 can have a corresponding visual representation (e.g., a particular shape and / or color, such as red). Additional user input (e.g., pressing the "Enter" key on a keyboard) can result in the generation of a segmentation using the particular drill cuttings (e.g., drill cuttings 122) and the negative point 124. Fig.10 The image 126 of the present invention may be used to annotate the image 126. For example, in conjunction with block 90 of method 78, the image annotations described may be used for interpretation of the image (e.g., by an operator at a well site) and for classifying different rock particles into different lithologies. Additionally, in some embodiments, the annotations may be used to improve the segmentation model applied by the prediction engine 74 utilizing the LFM.

[0058] It is contemplated that the user may make other annotations in conjunction with the image to be processed via the prediction engine 74 utilizing the LFM. For example, in response to user input, such as pressing the "Backspace" key, all points on the image generated by the GUI may be removed. Annotations and subsequent segmentation adjustments may continue until the user is satisfied with the results of the operation. This may allow the user to interact to optimally detect rock grains. In addition, the GUI described herein allows for interactive modification (e.g., correction) of the segmentation with minimal user interaction (i.e., a reduced number of mouse clicks, keyboard inputs, etc.). This allows the resulting computer system (i.e., the analysis system 60) to improve the efficiency with which the user can navigate through the various views and windows by providing instant or near instant confirmation of the amount of work performed and / or to be performed, by using the described visual overlay techniques implemented via the GUI. In effect, by providing an overlay, efficient use of processing power, memory, storage space, network bandwidth, and / or other computing resources is achieved.

[0059] Techniques and systems disclosed herein relate to utilizing image analysis via a prediction engine 74 utilizing LFM. Additional aspects relate to using a GUI to allow a user to interact with an image of rock particles, which can then be segmented using the prediction engine 74 utilizing LFM. The results can be used as input to allow control of related equipment, such as the drilling system 12 and / or a drilling plan for the drilling system 12, based on the lithology of the rock sample 47 determined from the segmented image generated by the prediction engine 74 utilizing LFM. Although the examples described above are illustrated with respect to wellbores on land, similar methods can be applied to any acquisition configuration.

[0060] For the purpose of explanation, the foregoing description has been described with reference to specific embodiments. However, the above illustrative discussion is not intended to be exhaustive or limit the present disclosure to the precise form disclosed. In view of the above teachings, many modifications and variations are possible. In addition, the order in which the elements of the method described herein are illustrated and described can be rearranged, and / or two or more elements can appear at the same time. The embodiments are selected and described in order to best explain the principles of the present disclosure and its practical application, so that other technical personnel in the art can best utilize the present disclosure and various embodiments with various modifications to be suitable for the intended specific purpose.

[0061] The technology presented and claimed herein is cited and applied to substantial objects and specific examples of a practical nature that significantly advance the art and are therefore not abstract, intangible, or purely theoretical. In addition, if any claim appended at the end of this specification contains one or more elements designated as "means for [performing] ... [function]" or "steps for [performing] ... [function]," then those elements should be interpreted under 35 U.S.C. § 112(f). However, for any claim containing an element designated in any other manner, it is intended that such element should not be interpreted under 35 U.S.C. § 112(f).

Claims

1. A system comprising: one or more processors; and A memory, accessible by one or more processors, storing instructions that, when executed by the one or more processors, cause the one or more processors to perform operations, including: receiving image data of an image of a rock sample from an imaging system; analyzing the image data via an artificial intelligence (AI) model trained using a plurality of images; and A segmented image is generated based on analysis of the image data via the AI ​​model, wherein the segmented image includes a representation of the rock sample of the image data. 2 . The system of claim 1 , wherein analyzing the image data further comprises applying a first level of resolution to the image data when generating the segmented image.

3. The system of claim 2, wherein the first level of resolution corresponds to one or a combination of a default or user-selected resolution setting.

4. The system of claim 2, wherein analyzing the image data further comprises applying a second level of resolution to the image data in place of the first level of resolution when generating the segmented image.

5. The system according to claim 2, further comprising: receiving a user input corresponding to a first region of the segmented image; and An adjusted segmented image is generated by removing, from the segmented image, rock samples associated with the first region of the segmented image.

6. The system according to claim 5, further comprising: Segmented images adjusted through AI model analysis; and Based on analysis of the adjusted segmented image via the AI ​​model, a second segmented image is generated, wherein the second segmented image includes a second representation of the rock sample of the image data. 7 . The system of claim 6 , wherein analyzing the adjusted segmented image further comprises applying a second level of resolution to the adjusted segmented image to generate a second segmented image.

8. A computer-implemented method comprising: receiving image data of an image of a rock sample from an imaging system; analyzing the image data via an artificial intelligence (AI) model trained using the plurality of images; and A segmented image is generated based on analysis of the image data via the AI ​​model, wherein the segmented image includes a representation of the rock sample of the image data.

9. The computer-implemented method of claim 8, further comprising: Analyzing the segmented image via an AI model; and Based on analysis of the segmented image via the AI ​​model, a second segmented image is generated, wherein the second segmented image includes a second representation of the rock sample of the image data.

10. The computer-implemented method of claim 9, wherein analyzing the segmented image further comprises applying a second level of resolution to the segmented image to generate a second segmented image.

11. A device comprising: monitor; a processor communicatively coupled to the display; and A memory communicatively coupled to the processor, the memory storing instructions that, when executed, cause the processor to perform operations comprising: Generate a graphical user interface (GUI); generating an image of the rock sample corresponding to the image data received from the imaging system for presentation on a display; receiving a first input via a first user interaction with the GUI; generating a visual icon for display on the display at a specific location on an image of the rock sample displayed on the display, wherein the specific location is determined based on the first input; receiving a second input via a second user interaction with the GUI; and In response to a second input, at least a portion of the image data is analyzed via an artificial intelligence (AI) model trained using multiple images, and a segmented image is generated based on the analysis of at least a portion of the image data by the AI ​​model, wherein the segmented image includes a representation of at least a portion of the rock sample.

12. The apparatus of claim 11, wherein at least a portion of the image data analyzed by the AI ​​model corresponds to a specific location on the image of the rock sample.

13. The apparatus according to claim 11, further comprising: receiving a second input via a second user interaction with the GUI; and A second visual icon displayed on the display is generated at a second specific location on the image of the rock sample displayed on the display, wherein the second specific location is determined based on the second input.

Images