Fracture automatic identification method based on three-dimensional fitting borehole imaging logging

Abstract

The application discloses a wellbore imaging logging fracture automatic identification method based on three-dimensional fitting, a processor and a machine readable storage medium, and the identification method comprises the following steps: acquiring a two-dimensional wellbore imaging logging image; inputting the two-dimensional wellbore imaging logging image into a preset deep learning image segmentation algorithm to obtain an extraction result of a fracture development area; obtaining a three-dimensional projection position of the fracture development area according to a caliper value and the extraction result; extracting spatial points in the three-dimensional projection position that are connected to each other as independent fitting areas; selecting a plurality of independent fitting areas in the three-dimensional projection position and performing plane fitting on the independent fitting areas respectively to obtain a plurality of initial correlation coefficients; determining an ideal fitting surface according to the plurality of initial correlation coefficients; and determining a three-dimensional wellbore imaging logging image according to the ideal fitting surface. Through the technical scheme, the influence of caliper variation on the fracture response characteristics of wellbore imaging logging can be eliminated, the information of the fracture can be restored more accurately, and the interpretation accuracy of wellbore imaging logging fracture is improved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of oil and gas exploration, and particularly relates to a wellbore imaging logging fracture automatic identification method based on three-dimensional fitting, a processor and a machine readable storage medium. BACKGROUND

[0002] The wellbore imaging logging technology can provide 360° full-range petrophysical two-dimensional image information of the wellbore wall at high resolution. Since the fracture surface of the structure is relatively flat, the physical properties (resistivity, acoustic time difference) of the fracture surface are different from those of the background area, so the fracture surface will be obviously displayed on the wellbore imaging logging image. Meanwhile, the fracture surface is mostly manifested as a typical sinusoidal curve on the two-dimensional wellbore imaging logging image. At present, the criterion for interpreting the wellbore imaging logging fracture is to find the sinusoidal curve.

[0003] The conventional wellbore imaging logging fracture identification is mainly based on manual interpretation, which requires a higher experience of the interpreter. With the progress of computer technology, the technology of automatically identifying the fracture by using artificial intelligence method is also developing. The current wellbore imaging logging fracture automatic identification method is mainly image processing, which finds the sinusoidal curve on the two-dimensional unfolded wellbore imaging logging image. Due to the difference in the physical properties of the strata, the phenomenon of diameter expansion or reduction will inevitably occur during drilling. When the well diameter changes, the fracture response characteristics in the wellbore imaging logging will also be affected, and no longer present a typical sinusoidal curve. When the well diameter changes, the sinusoidal curve on the two-dimensional wellbore imaging logging curve will become very irregular, making it difficult to interpret the wellbore imaging logging fracture according to the criterion of finding the sinusoidal curve, which seriously affects the accuracy of the wellbore imaging logging fracture interpretation. SUMMARY

[0004] In view of the above defects or deficiencies of the prior art, the present application provides a wellbore imaging logging fracture automatic identification method based on three-dimensional fitting, a processor and a machine readable storage medium, which aims to solve the technical problem that when the well diameter changes, the interpretation of the wellbore imaging logging fracture according to the criterion of finding the sinusoidal curve will seriously affect the accuracy of the wellbore imaging logging fracture interpretation.

[0005] In order to achieve the above-mentioned purpose, the present application provides a wellbore imaging logging fracture automatic identification method based on three-dimensional fitting, wherein the wellbore imaging logging fracture automatic identification method based on three-dimensional fitting comprises:

[0006] obtaining a two-dimensional wellbore imaging logging image;

[0007] inputting the two-dimensional wellbore imaging logging image into a preset deep learning image segmentation algorithm to obtain an extraction result of a fracture development area;

[0008] obtaining a three-dimensional projection position of the fracture development area according to the well diameter value and the extraction result;

[0009] extracting spatial points in the three-dimensional projection position as independent fitting regions;

[0010] selecting a plurality of independent fitting regions in the three-dimensional projection position to perform plane fitting respectively to obtain a plurality of initial correlation coefficients;

[0011] determining an ideal fitting plane according to the plurality of initial correlation coefficients;

[0012] determining a three-dimensional borehole imaging logging image according to the ideal fitting plane.

[0013] In the embodiment of the present application, the method for determining the ideal fitting plane according to the plurality of initial correlation coefficients comprises:

[0014] determining an initial fitting plane corresponding to the initial correlation coefficient greater than the preset coefficient threshold as a candidate fitting plane;

[0015] adding the remaining independent fitting regions in the candidate fitting plane to expand the spatial points to obtain an expanded fitting plane and calculate an expanded correlation coefficient;

[0016] determining the ideal fitting plane according to the expanded correlation coefficient.

[0017] In the embodiment of the present application, the method for determining the ideal fitting plane according to the expanded correlation coefficient comprises:

[0018] calculating an evaluation index of the expanded fitting plane according to the expanded correlation coefficient and the number of spatial points in the corresponding expanded fitting plane;

[0019] determining the ideal fitting plane according to the evaluation index.

[0020] In the embodiment of the present application, the method for obtaining the three-dimensional projection position of the fracture development region according to the caliper value and the extraction result comprises:

[0021] performing filtering processing on the extraction result to obtain a filtering result of the fracture development region;

[0022] performing three-dimensional space projection on the filtering result combined with the caliper value to obtain the three-dimensional projection position of the fracture development region.

[0023] In the embodiment of the present application, the method for generating the preset deep learning image segmentation algorithm comprises:

[0024] designing a neural network structure;

[0025] obtaining a manually labeled borehole imaging logging training image;

[0026] performing image preprocessing on the borehole imaging logging training image;

[0027] The trained image of the well wall after image preprocessing is input into the designed neural network for training, and the parameters of the neural network are updated after the training is completed.

[0028] When the error of the training output result is less than the preset error value, the preset deep learning image segmentation algorithm is determined according to the current updated parameters.

[0029] In the embodiment of the present application, the number of ideal fitting planes is at least two, and the determination method of the at least two ideal fitting planes is:

[0030] After the first ideal fitting plane is determined, the independent fitting region for determining the first ideal fitting plane is deleted in the three-dimensional projection position to obtain a cyclic projection position.

[0031] In the cyclic projection position, a plurality of independent fitting regions are continuously selected for plane fitting, and the next ideal fitting plane is determined according to the plane fitting result.

[0032] In the embodiment of the present application, the identification method further comprises:

[0033] The position parameters of the fracture development region are determined according to the ideal fitting plane.

[0034] In the embodiment of the present application, the position parameters include the dip angle and the dip direction, and the calculation formula of the dip angle is:

[0035]

[0036] The calculation formula of the dip direction is:

[0037] Or

[0038]

[0039] Wherein, the values A, B and C are the normal vectors (A, B, C) of the ideal fitting plane.

[0040] To achieve the above-mentioned purpose, the second aspect of the present application provides a processor configured to execute the three-dimensional fitting based wellbore imaging logging fracture automatic identification method described above.

[0041] To achieve the above-mentioned purpose, the third aspect of the present application provides a machine readable storage medium, and the machine readable storage medium stores instructions, and the instructions are executed by the processor to realize the three-dimensional fitting based wellbore imaging logging fracture automatic identification method described above.

[0042] Through the above technical solution, the three-dimensional fitting based wellbore imaging logging fracture automatic identification method provided by the embodiment of the present application has the following beneficial effects:

[0043] In the technical solution, the fracture development area of a two-dimensional borehole imaging logging image is extracted by an image segmentation algorithm, the three-dimensional projection position of the fracture development area is restored according to a caliper value and the extraction result, the ideal fitting plane of the fracture development area is determined by using the connectivity of a space point, plane fitting and correlation analysis, and finally the three-dimensional borehole imaging logging image of the fracture development area can be determined according to the ideal fitting plane, so that the influence of caliper variation on the fracture response characteristics of the borehole imaging logging can be eliminated, the information of the fracture can be restored more accurately, and the fracture interpretation accuracy of the borehole imaging logging is improved.

[0044] Other features and advantages of the embodiments of the present application will be described in detail in the following specific implementation part. BRIEF DESCRIPTION OF DRAWINGS

[0045] The accompanying drawings are included to provide a further understanding of the embodiments of the present application, and constitute a part of the specification, and are used together with the following specific implementation to explain the embodiments of the present application, but do not constitute a limitation on the embodiments of the present application. In the drawings:

[0046] Figure 1 It is a flow chart of a three-dimensional fitting based borehole imaging logging fracture automatic identification method according to an embodiment of the present application;

[0047] Figure 2 It is a flow chart of step 600 in the three-dimensional fitting based borehole imaging logging fracture automatic identification method according to an embodiment of the present application;

[0048] Figure 3 It is a flow chart of step 630 in the three-dimensional fitting based borehole imaging logging fracture automatic identification method according to an embodiment of the present application;

[0049] Figure 4 It is a derivation schematic diagram of a position parameter calculation formula of an ideal fitting plane in an embodiment of the present application;

[0050] Figure 5 It is a recognition result schematic diagram of three-dimensional fitting according to the recognition method of the present application. DETAILED DESCRIPTION

[0051] The specific implementation of the embodiments of the present application will be described in detail below in combination with the drawings. It should be understood that the specific implementation described herein is only used to illustrate and explain the embodiments of the present application, and is not used to limit the embodiments of the present application.

[0052] It should be noted that if the application embodiments involve directionality indications (such as up, down, left, right, front, back, etc.), the directionality indications are only used to explain the relative position relationship, motion condition, etc. between components in a certain specific posture (as shown in the drawings), and if the specific posture changes, the directionality indications also change accordingly.

[0053] In addition, if the application embodiments involve descriptions such as "first", "second", etc., the descriptions of "first", "second", etc. are only for description purposes and cannot be understood as indicating or implying the relative importance or implicitly indicating the number of indicated technical features. Therefore, the features defined as "first", "second" can explicitly or implicitly include at least one of the features. In addition, the technical solutions of each embodiment can be combined with each other, but it must be based on the realization of a person skilled in the art, and when the combination of technical solutions contradicts each other or cannot be realized, it should be considered that the combination of technical solutions does not exist and is not within the protection scope required by the application.

[0054] Figure 1 The flowchart of the fracture automatic identification method based on three-dimensional fitting of the borehole imaging logging according to an embodiment of the application is schematically shown. As shown in Figure 1 The application provides a fracture automatic identification method based on three-dimensional fitting of borehole imaging logging, wherein the fracture automatic identification method based on three-dimensional fitting of borehole imaging logging comprises the following steps:

[0055] Step 100, acquiring a two-dimensional borehole imaging logging image.

[0056] Specifically, the two-dimensional borehole imaging logging image can be measured by a logging instrument.

[0057] Step 200, inputting the two-dimensional borehole imaging logging image into a preset deep learning image segmentation algorithm to obtain an extraction result of a fracture development area.

[0058] Further, the two-dimensional borehole imaging logging image can be input into the preset deep learning image segmentation algorithm, and a sub-image containing fracture information can be obtained through calculation, and the sub-image containing fracture information is the extraction result.

[0059] Step 300, obtaining a three-dimensional projection position of the fracture development area according to the caliper value and the extraction result.

[0060] Specifically, according to the extraction result of the sub-image containing fracture information and the caliper value, the borehole fracture position in the two-dimensional image can be restored to the three-dimensional space to obtain the three-dimensional projection position of the fracture development area.

[0061] Step 400, extract the spatial points in the three-dimensional projection position as an independent fitting region.

[0062] Specifically, the Matlab image processing function regionprops can be used as a preset image region algorithm, and the spatial points in the three-dimensional projection position are divided into regions by the preset image region algorithm, so that the divided region is the independent fitting region of the three-dimensional projection position as long as the spatial points in the divided region are mutually connected or single spatial points.

[0063] Step 500, select a plurality of independent fitting regions in the three-dimensional projection position to perform plane fitting respectively to obtain a plurality of initial correlation coefficients.

[0064] Further, for the three-dimensional projection position of the fracture development region, the three-dimensional projection position can be divided into a plurality of independent fitting regions according to the connectivity of the spatial points, the spatial points in the independent fitting region are mutually connected, the plurality of independent fitting regions are sorted according to the number of spatial points, and the independent fitting region with less than 3 spatial points is deleted, and then the correlation coefficient of the spatial points in each independent fitting region is calculated, in addition, the first 6 independent fitting regions can be selected for plane fitting as the initial fitting plane, and the correlation coefficient of the independent fitting region itself is used as the initial correlation coefficient of the initial fitting plane.

[0065] Step 600, determine the ideal fitting plane according to the plurality of initial correlation coefficients.

[0066] Specifically, the plurality of independent fitting regions can obtain a plurality of initial correlation coefficients through plane fitting, and then the initial correlation coefficient with the highest value can be selected as the corresponding fitting plane to determine the ideal fitting plane by comparing the plurality of initial correlation coefficients.

[0067] Step 700, determine the three-dimensional borehole imaging logging image according to the ideal fitting plane.

[0068] In the above technical solution, the fracture development region of the two-dimensional borehole imaging logging image is extracted by the image segmentation algorithm, the three-dimensional projection position of the fracture development region is restored according to the caliper value and the extraction result, the ideal fitting plane of each fracture development region is determined by using the connectivity of the spatial points, the plane fitting and the correlation analysis, and finally the three-dimensional borehole imaging logging image of the fracture development region can be determined according to the ideal fitting plane, so that the influence of the caliper change on the fracture response characteristics of the borehole imaging logging can be eliminated, the information of the fracture can be restored more accurately, and the fracture interpretation accuracy of the borehole imaging logging is improved.

[0069] Figure 2 A flowchart of step 600 in the borehole imaging logging fracture automatic identification method based on three-dimensional fitting according to an embodiment of the present application is schematically shown. Figure 2As shown in the embodiment of the present application, step 600, determining the ideal fitting plane according to the plurality of initial correlation coefficients comprises:

[0070] Step 610, determining the initial fitting plane corresponding to the initial correlation coefficient greater than the preset coefficient threshold as the candidate fitting plane.

[0071] Specifically, the plurality of initial correlation coefficients obtained by plane fitting of the selected plurality of independent fitting regions are compared with the preset coefficient threshold respectively, if there are m (m≥2) initial correlation coefficients greater than the preset coefficient threshold, then the initial fitting plane corresponding to the m initial correlation coefficients is determined as the candidate fitting plane. More specifically, the preset coefficient threshold R t is greater than 0.7, and the correlation coefficient can be obtained by regression analysis.

[0072] Step 620, adding the remaining independent fitting regions in the candidate fitting plane for spatial point expansion to obtain an expanded fitting plane, and calculating an expanded correlation coefficient.

[0073] In addition, after determining the m candidate fitting planes, the m candidate fitting planes can be continuously expanded for spatial points. Based on the spatial points in the independent fitting regions, other remaining independent fitting regions in the three-dimensional projection position are randomly added for spatial point fitting, if the correlation coefficient of the fitting plane is unchanged or increased compared with the initial correlation coefficient, the independent fitting region is added; if the correlation coefficient is reduced, the independent fitting region is discarded, until all the remaining independent fitting regions are added or discarded, so that the final expanded fitting plane can be obtained, and the expanded correlation coefficient is recorded.

[0074] Step 630, determining the ideal fitting plane according to the expanded correlation coefficient.

[0075] By selecting the initial fitting plane with a larger initial correlation coefficient as the candidate fitting plane, then expanding the spatial points of the candidate fitting plane, an expanded fitting plane containing more independent fitting regions can be obtained, so that the ideal fitting plane can be finally determined according to the expanded correlation coefficient corresponding to the expanded fitting plane, so as to ensure the accuracy of the three-dimensional projection position of the fracture development region.

[0076] Figure 3 The flowchart of step 630 in the three-dimensional fitting based borehole imaging logging fracture automatic identification method according to an embodiment of the present application is schematically shown as Figure 3 As shown in the embodiment of the present application, step 630, determining the ideal fitting plane according to the expanded correlation coefficient comprises:

[0077] Step 640, calculating the evaluation index of the expanded fitting plane according to the expanded correlation coefficient and the number of spatial points in the corresponding expanded fitting plane.

[0078] Step 650, determining the ideal fitting plane according to the evaluation index.

[0079] Further, the evaluation index of each of the m extended fitting planes can be calculated by substituting the extended correlation coefficient and the number of all space points in the extended fitting plane into the evaluation index calculation formula, and the extended fitting plane with the maximum evaluation index is determined as the ideal fitting plane, so that the reliability of the ideal fitting plane can be ensured.

[0080]

[0081] wherein, R i is the extended correlation coefficient of the extended fitting plane, is the sum of the extended correlation coefficients of all the extended fitting planes, k i is the number of the space points in the extended fitting plane, is the sum of the numbers of the space points in all the extended fitting planes, i=1, 2, …, m, j=1, 2, …, m.

[0082] According to the above evaluation index calculation formula, the m extended fitting planes can be evaluated, and the plane with the maximum evaluation index is selected as the ideal fitting plane.

[0083] In the embodiment of the present application, step 300, the three-dimensional projection position of the fracture development region is obtained according to the caliper value and the extraction result, which comprises:

[0084] Step 310, filtering the extraction result to obtain the filtering result of the fracture development region.

[0085] Specifically, the extraction result is filtered by using a median filtering method or the like to reduce the influence of noise on the extraction result.

[0086] Step 320, combining the filtering result with the caliper value to perform three-dimensional space projection to obtain the three-dimensional projection position of the fracture development region.

[0087] Further, the caliper value at the depth of the fracture development region can be measured as r z , the coordinates (x z , y z , z) of the well axis center at the depth of the fracture development region can be calculated according to the filtering result, and the two-dimensional image included angle θ of the fracture development region, so that the coordinates of the extracted well wall fracture position in the two-dimensional image projected to the three-dimensional space are (x z +r z cosθ, y z +r z sinθ, z).

[0088] In the embodiment of the present application, the generation method of the preset deep learning image segmentation algorithm comprises:

[0089] Step 210, designing a neural network structure.

[0090] Step 220, obtaining manually labeled borehole imaging logging training images.

[0091] Step 230, image preprocessing the borehole imaging logging training images.

[0092] Step 240, inputting the borehole imaging logging training images after image preprocessing into the designed neural network for training, and updating the parameters of the neural network after the training is completed.

[0093] Step 250, when the error of the training output result is less than a preset error value, determining the preset deep learning image segmentation algorithm according to the current updated parameters.

[0094] In order to obtain a preset deep learning image segmentation algorithm which can facilitate the extraction of a sub-image containing fracture information in a two-dimensional borehole imaging logging image, the preset deep learning image segmentation algorithm can be established. First, a neural network structure is designed, and an algorithm is programmed and implemented, mainly including an input layer, a convolution layer, an activation layer, a pooling layer and an output layer. The convolution layer is used for feature extraction, and the pooling layer is used to highlight main features and reduce the size of the image and the amount of calculation. Then, a certain number of manually labeled borehole imaging logging training images are selected, and the borehole imaging logging training images are inputted for image preprocessing, which are stretched or compressed into pictures of the same size. A plurality of pictures are inputted into the neural network for training as a group, and the parameters are updated once after the training of a group of pictures is completed. When the error of the training output result is less than a preset error value, the preset deep learning image segmentation algorithm can be determined according to the current updated parameters. Thus, after a two-dimensional borehole imaging logging image is inputted into the preset deep learning image segmentation algorithm, an extraction result containing fracture information can be outputted.

[0095] In the embodiment of the present application, the activation function of the preset deep learning image segmentation algorithm is a linear rectifier function. Specifically, the formula of the linear rectifier function is f(x) = max(0, x).

[0096] In the embodiment of the present application, the number of ideal fitting planes can be at least two, and the determination method of the at least two ideal fitting planes is as follows:

[0097] After the first ideal fitting plane is determined, the independent fitting region in which the first ideal fitting plane is determined is deleted in the three-dimensional projection position to obtain a cyclic projection position.

[0098] In the cyclic projection position, a plurality of independent fitting regions are continuously selected for plane fitting, and the next ideal fitting plane is determined according to the plane fitting result.

[0099] Specifically, after the first ideal fitting plane is determined, all the independent fitting regions in which the first ideal fitting plane is determined can be deleted from the three-dimensional projection position, and the remaining independent fitting regions can be taken as the cyclic projection position, and the cyclic projection position can replace the three-dimensional projection position in step 500 and step 500 and step 600 can be repeated to obtain the next ideal fitting plane. In addition, the above operation can be repeated to continue deleting the independent fitting region in which the last ideal fitting plane is determined from the last cyclic projection position, and the remaining independent fitting region can be taken as the next cyclic projection position, and the termination condition of the cycle is that the number of the remaining independent fitting regions is less than p (p≥6), at this time, it can be considered that the fractures in the wellbore imaging logging image are extracted, and at least two ideal fitting planes can constitute the final three-dimensional wellbore imaging logging image. More specifically, in the embodiment, the number of the independent fitting regions initially split according to the connectivity can be 49.

[0100] In the embodiment of the present application, the identification method further comprises:

[0101] Step 800, determining the position parameters of the fracture development region according to the ideal fitting plane.

[0102] That is, after obtaining the three-dimensional wellbore imaging logging image, it is not necessary to rely on experienced personnel for manual interpretation, and the information of the fractures can be restored more intelligently to further achieve the purpose of improving the fracture interpretation accuracy of the wellbore imaging logging.

[0103] In the embodiment of the present application, the position parameters include the dip angle and the dip direction, and the calculation formula of the dip angle is:

[0104]

[0105] The calculation formula of the dip direction is:

[0106] or

[0107]

[0108] Wherein, the values of A, B and C are the normal vector (A, B, C) of the ideal fitting plane.

[0109] Specifically, the derivation steps of the calculation formula of the dip angle and the dip direction are as follows:

[0110] (1) The plane calculation formula of the ideal fitting plane is Ax+By+Cz+D=0, and the normal vector of the ideal fitting plane is (A, B, C). The normal vector of the horizontal plane is (0, 0, 1), that is, the Z-axis direction, which can be specifically referred to as . Figure 4

[0111] ​(2) The inclination angle a (0°≤a≤90°) of the fracture surface is the included angle (or its supplementary angle) between the normal vector of the fracture surface and the normal vector of the horizontal plane The calculation formula of the inclination angle a can be converted from the vector dot product formula.

[0112] (3) Let the vector be the normal vector of the fracture surface projected on the bottom surface, then the vector is (A, B, 0), and the included angle between the vector and the positive unit vector of the y-axis is the strike angle β (0°≤β≤360°) of the fracture surface, and the unit vector is (0, 1, 0), so the following can be derived:

[0113]

[0114] The value range of β obtained by the above formula is [0, 180°], β can be converted to the value range [0, 360°], then when A≥0, otherwise, The strike of the fracture surface can be calculated by adding or subtracting 90° to the strike angle.

[0115] Specifically, the generation steps of the three-dimensional borehole wall imaging logging image provided by the present application can be:

[0116] Firstly, a two-dimensional borehole wall imaging logging image can be obtained by a logging instrument;

[0117] Secondly, the two-dimensional borehole wall imaging logging image is input into a preset deep learning image segmentation algorithm to obtain an extraction result of a fracture development area;

[0118] Thirdly, the extraction result is filtered to obtain a filtering result of the fracture development area;

[0119] Fourthly, the filtering result is combined with a well diameter value to project in a three-dimensional space to obtain a three-dimensional projection position of the fracture development area;

[0120] Fifthly, spatial points connected to each other in the three-dimensional projection position of the fracture development area are extracted as independent fitting areas;

[0121] Sixthly, a plurality of independent fitting areas are selected in the three-dimensional projection position and are respectively subjected to plane fitting to obtain a plurality of initial correlation coefficients;

[0122] Seventhly, an initial fitting plane corresponding to an initial correlation coefficient greater than a preset coefficient threshold value is determined as a candidate fitting plane;

[0123] Step 8: Add the remaining independent fitting regions to the candidate fitting surface to expand the spatial points, so as to obtain the expanded fitting surface and calculate the expanded correlation coefficient.

[0124] Step 9: Calculate the evaluation index for the expanded fitted surface based on the expanded correlation coefficient;

[0125] Step 10: Determine the ideal fitting surface based on the evaluation indicators;

[0126] Step 11: Obtain the three-dimensional wellbore imaging logging image based on the ideal fitting surface.

[0127] Furthermore, taking a two-dimensional wellbore imaging logging image of a certain well as an example, intelligent fracture identification was performed. The identification method of this invention was used for three-dimensional plane fitting, and the identification results are as follows: Figure 5 As shown.

[0128] The original wellbore imaging logging image is processed by a deep learning image segmentation algorithm to obtain a sub-image containing fracture information. This sub-image is then denoised. The remaining discontinuous fracture segments, combined with wellbore data, are projected into a 3D space. Correlation analysis is then performed on these fracture segments, and the fracture regions are combined using iterative fitting principles to obtain the fracture surface in 3D space. The fracture orientation is then determined based on the parameters of the fracture surface, such as... Figure 5 The two fractures in the image were extracted, with dip angles of 47.16° and 46.1° from top to bottom, and dip directions of 268.6° and 265.3°, respectively. Compared to traditional two-dimensional wellbore imaging logging fracture identification, this invention can eliminate the influence of wellbore diameter variations on fracture identification. Even if the fracture response in the wellbore imaging logging image is not a sinusoidal curve, it can still achieve automatic fracture identification in three-dimensional space. This method improves the accuracy of automatic fracture identification in wellbore imaging logging, intuitively and clearly displays the distribution morphology of fractures in the formation, and provides technical support for parameter extraction and fine characterization of subsurface fractures.

[0129] Furthermore, in another embodiment of the present invention, a processor is provided, configured to execute the above-described method for automatic identification of fractures in wellbore imaging logging based on three-dimensional fitting.

[0130] In another embodiment of the present invention, a machine-readable storage medium is provided, on which instructions are stored. When the instructions are executed by a processor, they implement the above-described method for automatic identification of fractures in wellbore imaging logging based on three-dimensional fitting.

[0131] Those skilled in the art will appreciate that embodiments of the application can be readily used as software, hardware, or a combination of software and hardware. In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0132] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more functions specified in the flowchart block or blocks. Figure 1 means for functionally implementing one or more functions specified in the flowchart block or blocks.

[0133] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the function specified in the flowchart block or blocks. Figure 1 one or more functions specified in the flowchart block or blocks. Figure 1 means for functionally implementing one or more functions specified in the flowchart block or blocks.

[0134] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more functions specified in the flowchart block or blocks. Figure 1 means for functionally implementing one or more functions specified in the flowchart block or blocks.

[0135] In one typical configuration, the computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0136] The memory can include non-persistent memory and / or persistent memory, such as flash memory, read-only memory (ROM), and / or volatile / non-volatile random access memory (RAM), among others. The memory is an example of computer readable media.

[0137] Computer-readable media includes permanent and non-permanent, movable and non-movable media that can be implemented by any method or technology to store information. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information accessible to a computing device. According to the definition herein, computer-readable media does not include transitory media such as modulated data signals and carriers.

[0138] It should also be noted that the terms "comprising", "containing", or any other variant thereof are intended to cover non-exclusive inclusions, so that a process, method, article or apparatus that includes a list of elements does not only include those elements, but also includes other elements not explicitly listed, or further includes elements inherent in such a process, method, article or apparatus. Without more limitations, the element defined by the statement "comprising a" does not exclude the presence of additional identical elements in the process, method, article or apparatus that includes the element.

[0139] The above only is an embodiment of the present application, and is not used to limit the present application. For those skilled in the art, the present application can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application shall be included in the scope of claims of the present application.

Claims

1. A three-dimensional fitting-based automatic fracture identification method for borehole wall imaging logging, characterized in that, The identification method comprises: obtaining a two-dimensional borehole imaging logging image; inputting the two-dimensional borehole imaging logging image into a preset deep learning image segmentation algorithm to obtain an extraction result of a fracture development region; obtaining a three-dimensional projection position of the fracture development region according to a caliper value and the extraction result; extracting spatial points in the three-dimensional projection position that are connected to each other as independent fitting regions; selecting a plurality of the independent fitting regions in the three-dimensional projection position to perform plane fitting respectively to obtain a plurality of initial correlation coefficients; determining an ideal fitting surface according to the plurality of initial correlation coefficients; determining a three-dimensional borehole imaging logging image according to the ideal fitting surface; the method of determining the ideal fitting surface according to the plurality of initial correlation coefficients comprises: determining an initial fitting surface corresponding to the initial correlation coefficient greater than a preset coefficient threshold as a candidate fitting surface; adding the remaining independent fitting regions in the candidate fitting surface to perform spatial point expansion to obtain an expanded fitting surface and calculate an expanded correlation coefficient; determining the ideal fitting surface according to the expanded correlation coefficient; the method of determining the ideal fitting surface according to the expanded correlation coefficient comprises: calculating an evaluation index of the expanded fitting surface according to the expanded correlation coefficient and the number of spatial points in the expanded fitting surface; determining the ideal fitting surface according to the evaluation index; the method of determining the ideal fitting surface according to the evaluation index comprises: substituting the expanded correlation coefficient and the number of all spatial points in the corresponding expanded fitting surface into an evaluation index calculation formula to calculate the evaluation index of the m expanded fitting surfaces respectively, and determining the expanded fitting surface with the maximum evaluation index calculation result as the ideal fitting surface, the evaluation index calculation formula being: ； wherein, is the expansion correlation coefficient for a certain expansion fitting surface, is the sum of the expansion correlation coefficients for all expansion fitting surfaces, is the number of scatter points within a certain expansion fitting surface, is the sum of the number of scatter points within all expansion fitting surfaces, .

2. The method of claim 1, wherein, the method of obtaining the three-dimensional projection position of the fracture development region according to the caliper value and the extraction result comprises: performing filtering processing on the extraction result to obtain a filtering result of the fracture development region; performing three-dimensional space projection on the filtering result combined with the caliper value to obtain the three-dimensional projection position of the fracture development region.

3. The method of claim 1, wherein, The method of generating the preset deep learning image segmentation algorithm comprises: designing a neural network structure; obtaining manually labeled borehole imaging logging training images; performing image preprocessing on the borehole imaging logging training images; inputting the borehole imaging logging training images after image preprocessing into the designed neural network for training, and updating parameters of the neural network after training is completed; when an error of a training output result is less than a preset error value, determining the preset deep learning image segmentation algorithm according to the current updated parameters.

4. The method of claim 1, wherein, The number of the ideal fitting surfaces is at least two, and the method of determining the at least two ideal fitting surfaces comprises: after determining a first ideal fitting surface, deleting the independent fitting region for determining the first ideal fitting surface in the three-dimensional projection position to obtain a circulating projection position; continuing to select a plurality of the independent fitting regions in the circulating projection position to perform plane fitting respectively, and determining a next ideal fitting surface according to a plane fitting result.

5. The method of claim 1 to 4, wherein, The identification method further comprises: determining a position parameter of the fracture development region according to the ideal fitting surface.

6. The method of claim 5, wherein, The position parameters include an inclination angle and a dip, and the inclination angle is calculated according to the following formula: ； The dip is calculated according to the following formula: (A > 0); or ，（A＜0）； Wherein, the values of A, B and C are normal vectors (A, B, C) of the ideal fitting surface.

7. A processor, comprising: The method is configured to perform the three-dimensional fitting-based automatic fracture identification method for borehole imaging logging according to any one of claims 1 to 6.

8. A machine-readable storage medium having stored thereon instructions, the instructions comprising: The instructions are executed by the processor to implement the three-dimensional fitting-based automatic fracture identification method for borehole imaging logging according to any one of claims 1 to 6.

Images