Intelligent three-dimensional segmentation method for three-dimensional digital core aperture characteristics and related equipment

Abstract

The invention discloses a three-dimensional digital core aperture feature intelligent three-dimensional segmentation method and related equipment, and the method comprises the steps: firstly obtaining an aperture feature contour and region in a two-dimensional image slice in a mode of learning a network model, achieving high speed and high efficiency, and then carrying out the segmentation of the aperture feature contour and region in a two-dimensional image slice through a mode of mapping and information merging. According to the invention, the method achieves the three-dimensional segmentation and extraction of the hole seam features of the whole three-dimensional digital rock core, achieves the automatic segmentation through the deep learning technology, is high in efficiency, is high in precision, is easy to popularize, can effectively solve the technical problem of the conventional three-dimensional image processing, and meets the demands of the related fields for the high-quality three-dimensional feature extraction. Compared with an artificial three-dimensional segmentation method, the method has a wide application prospect. The method not only can be used as a powerful auxiliary means for conventional digital core processing work, but also provides an efficient three-dimensional image analysis tool for microcosmic feature research in related research fields, and has important application value.

Description

Technical Field

[0001] The invention belongs to the field of three-dimensional digital core image processing, and specifically relates to a three-dimensional digital core pore feature intelligent stereoscopic segmentation method and related equipment. Background Art

[0002] 3D digital core is one of the commonly used data expression forms in rock physics research and oil and gas engineering development. It uses imaging technologies such as CT scanning, ultrasonic scanning or magnetic resonance scanning of physical cores to obtain digital expression of the 3D structural information inside the rock. With the help of 3D digital core technology, the microscopic structural characteristics and spatial distribution laws inside the rock can be accurately understood non-destructively.

[0003] In order to study the microstructure of different rock types and their impact on physical properties, it is necessary to extract and reconstruct key microstructural features, such as pore networks and particle size distribution, from three-dimensional digital core data sets. This requires three-dimensional image segmentation to segment different target areas from complex three-dimensional spatial structures. Three-dimensional image segmentation has always been a difficult and hot topic in computer vision and image processing, providing key technical support for research in multiple fields.

[0004] At present, the image segmentation methods of 3D digital cores are mainly divided into two categories: one is 3D segmentation software based on manual interaction, and the other is automatic 3D segmentation method based on algorithm model. The former relies on professional image processing software and the interactive tools it provides, and image processing experts manually outline the 3D target area to be segmented. This method is inefficient and prone to misjudgment; the latter proposes various algorithm models to achieve fully automatic 3D segmentation, but most models are difficult to achieve the high accuracy and efficiency required for practical use, and have poor versatility. Summary of the invention

[0005] The present invention provides a three-dimensional digital core pore feature intelligent stereoscopic segmentation method and related equipment, which solves the problem that the existing three-dimensional digital core image segmentation method is low in efficiency and prone to misjudgment.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] The intelligent stereo segmentation method of pore and fracture features of three-dimensional digital core includes:

[0008] Acquire a three-dimensional digital core stereoscopic image, and cut the three-dimensional digital core stereoscopic image into multiple two-dimensional image slices;

[0009] Inputting multiple two-dimensional image slices into the trained core pore feature instance segmentation model to obtain the pore feature contours and regions in the multiple two-dimensional image slices;

[0010] The hole feature contours and regions on each slice are mapped back to the three-dimensional space to reconstruct three three-dimensional hole segmentation structures;

[0011] The segmentation result information of the pixel points where the three-dimensional pore and fracture features are located is merged to realize the segmentation and extraction of the pore and fracture features of the three-dimensional digital core.

[0012] Preferably, the method for acquiring the core pore feature instance segmentation model is:

[0013] Collect a large number of two-dimensional images of core pore and fracture features, mark the contour curves of all pore and fracture features in the images, and construct a sample set of two-dimensional images of core pore and fracture features;

[0014] The learning network model is used to train the core pore feature two-dimensional image sample set to generate a core pore feature instance segmentation model.

[0015] Preferably, the learning network model adopts the Mask R-CNN deep learning network model.

[0016] Preferably, batch image enhancement processing is performed on all images before the contour curves of all hole and seam features in the images are marked, and the marking software tool uses LabelMe, RectLabel, LableBox or PixelAnnotationTool.

[0017] Preferably, the three-dimensional digital core stereoscopic imaging is divided into a plurality of two-dimensional image slices according to the three coordinate axis directions of X, Y and Z respectively.

[0018] Preferably, the hole feature contour and area on each slice are mapped back to the three-dimensional space, specifically, each two-dimensional image slice is spliced ​​and mapped back to the three-dimensional space in the directions of the three coordinate axes of X, Y, and Z according to the positional relationship between multiple slices.

[0019] Preferably, the segmentation result information of the pixel points where the three three-dimensional hole features are located is merged specifically as follows:

[0020] The three three-dimensional images are merged, and each pixel in the merged three-dimensional image contains different hole feature information in three directions. The different hole feature information in these three directions is taken as a union, that is, as long as there is a hole feature in the three directions of the pixel, the pixel will be marked as the corresponding hole feature.

[0021] The intelligent three-dimensional segmentation method calculation system of three-dimensional digital core pore characteristics includes:

[0022] Slice module: used to obtain three-dimensional digital core stereoscopic imaging and slice the three-dimensional digital core stereoscopic imaging into multiple two-dimensional image slices;

[0023] Segmentation module: input multiple two-dimensional image slices into the trained core pore feature instance segmentation model to obtain the pore feature contours and areas in the multiple two-dimensional image slices;

[0024] Mapping module: maps the hole feature contours and areas on each slice back to three-dimensional space and reconstructs three-dimensional hole segmentation structures;

[0025] Merging module: Merge the segmentation result information of the pixel points where the three-dimensional pore and fracture features are located to realize the segmentation and extraction of the pore and fracture features of the three-dimensional digital core.

[0026] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the intelligent stereoscopic segmentation method for pore and fracture features of three-dimensional digital cores are implemented.

[0027] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the intelligent stereoscopic segmentation method for pore and fracture features of three-dimensional digital cores.

[0028] Compared with the prior art, the present invention has the following beneficial effects: The present invention provides a method for intelligent stereoscopic segmentation of pore and fracture features of three-dimensional digital cores. The method first obtains the contours and areas of pore and fracture features in two-dimensional image slices in the form of a learning network model, which is fast and efficient. Then, the stereoscopic segmentation and extraction of pore and fracture features of the entire three-dimensional digital core are achieved through mapping and information merging. The deep learning technology is used as a whole to achieve automatic segmentation, which is efficient, accurate, and easy to promote. It can effectively solve the technical difficulties of previous three-dimensional image processing and meet the needs of related fields for high-quality three-dimensional feature extraction. Compared with the manual three-dimensional segmentation method, the present invention has broad application prospects. It can not only serve as a powerful auxiliary means for conventional digital core processing, but also provide an efficient three-dimensional image analysis tool for conducting microscopic feature research in related research fields, which has important application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is a flow chart of the intelligent stereoscopic segmentation method for pore and fracture features of three-dimensional digital cores of the present invention;

[0030] Figure 2 This is a block diagram of the computing system of the three-dimensional digital core pore and fracture feature intelligent stereo segmentation method of the present invention. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0032] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0034] In the description of the embodiments of the present invention, it should be noted that if the terms "upper", "lower", "horizontal", "inner", etc. indicate an orientation or positional relationship based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship in which the product of the invention is usually placed when in use, it is only for the convenience of describing the present invention and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. In addition, the terms "first", "second", etc. are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0035] In addition, if the term "horizontal" appears, it does not mean that the component must be absolutely horizontal, but can be slightly tilted. For example, "horizontal" only means that its direction is more horizontal than "vertical", which does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0036] In the description of the embodiments of the present invention, it is also necessary to explain that, unless otherwise clearly specified and limited, the terms "set", "install", "connect", and "connect" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal connection of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0037] In order to enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0038] like Figure 1 As shown, the present invention provides a three-dimensional digital core pore feature intelligent stereo segmentation method, comprising:

[0039] S101 acquires a three-dimensional digital core stereoscopic image, and cuts the three-dimensional digital core stereoscopic image into a plurality of two-dimensional image slices;

[0040] S102 inputs the multiple two-dimensional image slices into the trained core pore feature instance segmentation model to obtain the pore feature contours and regions in the multiple two-dimensional image slices;

[0041] S103 maps the hole feature contour and area on each slice back to the three-dimensional space to reconstruct three three-dimensional hole segmentation structures;

[0042] S104 combines the segmentation result information of the pixel points where the three three-dimensional pore and fracture features are located, so as to realize the segmentation and extraction of the pore and fracture features of the three-dimensional digital core.

[0043] Overall, deep learning technology is used to achieve automatic segmentation, which is efficient, accurate, and easy to promote. It can effectively solve the technical problems of previous three-dimensional image processing and meet the needs of related fields for high-quality three-dimensional feature extraction. Compared with manual three-dimensional segmentation methods, the present invention has broad application prospects. It can not only serve as a powerful auxiliary means for conventional digital core processing, but also provide an efficient three-dimensional image analysis tool for conducting microscopic feature research in related research fields, which has important application value.

[0044] The method for obtaining the core pore feature instance segmentation model is as follows:

[0045] Collect a large number of two-dimensional images of core pore and fracture features, mark the contour curves of all pore and fracture features in the images, and construct a sample set of two-dimensional images of core pore and fracture features;

[0046] The learning network model is used to train the core pore feature two-dimensional image sample set to generate a core pore feature instance segmentation model.

[0047] The learning network model adopts the Mask R-CNN deep learning network model.

[0048] Before marking the contour curves of all hole and seam features in the image, all images are subjected to batch image enhancement processing, and the marking software tool used is LabelMe, RectLabel, LableBox or PixelAnnotationTool.

[0049] The three-dimensional digital core stereo imaging is divided into multiple two-dimensional image slices according to the three coordinate axes of X, Y, and Z.

[0050] The hole feature contours and areas on each slice are mapped back to three-dimensional space. Specifically, each two-dimensional image slice is spliced ​​and mapped back to three-dimensional space in the directions of the three coordinate axes of X, Y, and Z according to the positional relationship between multiple slices.

[0051] The segmentation result information of the pixel points where the three 3D hole features are located is merged as follows:

[0052] The three three-dimensional images are merged, and each pixel in the merged three-dimensional image contains different hole feature information in three directions. The different hole feature information in these three directions is taken as a union, that is, as long as there is a hole feature in the three directions of the pixel, the pixel will be marked as the corresponding hole feature.

[0053] Another embodiment of the present invention provides a three-dimensional digital core pore feature intelligent stereoscopic segmentation method, specifically:

[0054] Based on the problems in the existing technology, there is an urgent need for a new method for automatic segmentation of three-dimensional digital cores that can ensure segmentation accuracy and efficiency and has strong versatility. Only such technical means can reduce labor costs, realize the deep mining of key features in massive three-dimensional image data sets, and provide strong support for related research and applications. This is also the technical starting point and goal of this research work.

[0055] In summary, three-dimensional digital core image segmentation is one of the main technical bottlenecks that restrict the research and application progress in this field. How to achieve high-quality automatic segmentation of three-dimensional images is a key technical problem that needs to be solved in this field. The method of the present invention is proposed to solve this technical problem and realize the accurate and efficient extraction of key feature areas in three-dimensional digital cores.

[0056] The specific steps include:

[0057] a. Collect a large number of two-dimensional images of core pore and fracture features, and perform batch image enhancement processing on all images, including image denoising, smoothing, sharpening and other operations. Then manually mark the contour curves of all pore and fracture features in the image to construct a sample set of two-dimensional images of core pore and fracture features;

[0058] b. Use the Mask R-CNN deep learning network model to train the core pore feature two-dimensional image sample set to generate a core pore feature instance segmentation model file, which fully learns all the pore features in the image.

[0059] c. The three-dimensional digital core stereo image is cut into multiple two-dimensional image slices according to the set slice step size in the three coordinate axes of X, Y, and Z. All slice images are segmented and reasoned through the trained core pore feature instance segmentation model to extract the pore feature contours and areas in all images.

[0060] d. In the directions of the three coordinate axes of X, Y, and Z, the two-dimensional segmentation results of the pore and fracture features on each slice are mapped back to the three-dimensional space according to the positional relationship between the slices, and three three-dimensional pore and fracture segmentation structures are reconstructed. Finally, the segmentation result information of the pixel points where the three three-dimensional pore and fracture features are located is merged, thus realizing the three-dimensional segmentation and extraction of the pore and fracture features of the entire three-dimensional digital core.

[0061] In step a, the contour curves of the hole features of the image are manually annotated, and the available annotation software tools include LabelMe, RectLabel, LableBox, PixelAnnotationTool, etc.

[0062] In step c, the three-dimensional digital core stereoscopic image is divided into multiple two-dimensional image slices according to the three coordinate axes of X, Y, and Z, and three two-dimensional image slice sets are obtained. Assuming that the lengths of the three-dimensional digital core in the X, Y, and Z axes are x, y, and z pixels respectively, and the slice step length is 1 pixel length, then the two-dimensional image slice set in the X direction is expressed as PX = {PX1, PX2, ..., PX x-1 , PX x}, where PX1 is the 2D image slice of the first pixel in the X direction on the YZ plane. Similarly, PY={PY1,PY2,……,PY y}, PZ={PZ1, PZ2,…, PZ z If the slice step length is 2 pixels, then the 2D image slice set in the X direction is expressed as PX = {PX1, PX3, ..., PX x-2 , PX x}.

[0063] In step c, when the slice step length is 1 pixel, all slice images are segmented and inferred by the trained core pore feature instance segmentation model to generate a two-dimensional slice image result set with pore feature information. PX corresponds to the result set PX' = {PX1', PX2', ..., PX x-1 ', PX x'}, and then map each image in the result set PX' back to the three-dimensional space according to the order of the slice positions, and a three-dimensional digital core image with stereoscopic pore feature information in the X direction can be reconstructed. Similarly, the three-dimensional digital core image with pore feature information in the Y and Z directions can also be reconstructed. Finally, the three three-dimensional stereo images are merged, and each pixel point in the merged three-dimensional stereo image contains different pore feature information in three directions. The union of these three information is taken, that is, as long as there is a pore feature in the three directions of the pixel point, the pixel point is marked as the corresponding pore feature.

[0064] The present invention also provides a three-dimensional digital core pore feature intelligent stereoscopic segmentation method calculation system, comprising:

[0065] Slice module: used to obtain three-dimensional digital core stereoscopic imaging and slice the three-dimensional digital core stereoscopic imaging into multiple two-dimensional image slices;

[0066] Segmentation module: input multiple two-dimensional image slices into the trained core pore feature instance segmentation model to obtain the pore feature contours and areas in the multiple two-dimensional image slices;

[0067] Mapping module: maps the hole feature contours and areas on each slice back to three-dimensional space and reconstructs three-dimensional hole segmentation structures;

[0068] Merging module: Merge the segmentation result information of the pixel points where the three-dimensional pore and fracture features are located to realize the segmentation and extraction of the pore and fracture features of the three-dimensional digital core.

[0069] A terminal device is provided in one embodiment of the present invention. The terminal device of this embodiment includes: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps in the above-mentioned method embodiments are implemented. Alternatively, when the processor executes the computer program, the functions of the modules / units in the above-mentioned device embodiments are implemented.

[0070] The computer program may be divided into one or more modules / units, and the one or more modules / units are stored in the memory and executed by the processor to accomplish the present invention.

[0071] The terminal device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0072] The processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0073] The memory may be used to store the computer programs and / or modules, and the processor implements various functions of the terminal device by running or executing the computer programs and / or modules stored in the memory and calling the data stored in the memory.

[0074] If the module / unit integrated in the terminal device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0075] Although the embodiments of the present invention are described above in conjunction with the accompanying drawings, the present invention is not limited to the above-mentioned specific embodiments and application fields, and the above-mentioned specific embodiments are only illustrative and instructive, rather than restrictive. Under the guidance of the specification, a person skilled in the art can also make many forms without departing from the scope of protection of the claims of the present invention, all of which belong to the scope of protection of the present invention.

Claims

1. A three-dimensional digital core pore feature intelligent stereo segmentation method, characterized in that: include: Acquire a three-dimensional digital core stereoscopic image, and cut the three-dimensional digital core stereoscopic image into multiple two-dimensional image slices; Inputting multiple two-dimensional image slices into the trained core pore feature instance segmentation model to obtain the pore feature contours and regions in the multiple two-dimensional image slices; The hole feature contours and regions on each slice are mapped back to the three-dimensional space to reconstruct three three-dimensional hole segmentation structures; The segmentation result information of the pixel points where the three-dimensional pore and fracture features are located is merged to realize the segmentation and extraction of the pore and fracture features of the three-dimensional digital core.

2. The intelligent stereoscopic segmentation method of pore and fracture features of three-dimensional digital core according to claim 1 is characterized in that: The method for obtaining the core pore feature instance segmentation model is as follows: Collect a large number of two-dimensional images of core pore and fracture features, mark the contour curves of all pore and fracture features in the images, and construct a sample set of two-dimensional images of core pore and fracture features; The learning network model is used to train the core pore feature two-dimensional image sample set to generate a core pore feature instance segmentation model.

3. The intelligent stereoscopic segmentation method for pore and fracture features of three-dimensional digital core according to claim 2 is characterized in that: The learning network model adopts the Mask R-CNN deep learning network model.

4. The intelligent stereoscopic segmentation method of pore and fracture features of three-dimensional digital core according to claim 2 is characterized in that: Before marking the contour curves of all hole and seam features in the image, all images are subjected to batch image enhancement processing, and the marking software tool used is LabelMe, RectLabel, LableBox or PixelAnnotationTool.

5. The intelligent stereoscopic segmentation method of pore and fracture features of three-dimensional digital core according to claim 1 is characterized in that: The three-dimensional digital core stereoscopic imaging is divided into a plurality of two-dimensional image slices according to the three coordinate axis directions of X, Y and Z.

6. The intelligent stereoscopic segmentation method of pore and fracture features of three-dimensional digital core according to claim 1 is characterized in that: The hole feature contours and areas on each slice are mapped back to three-dimensional space. Specifically, each two-dimensional image slice is spliced ​​and mapped back to three-dimensional space in the directions of the three coordinate axes of X, Y, and Z according to the positional relationship between multiple slices.

7. The intelligent stereoscopic segmentation method of pore and fracture features of three-dimensional digital core according to claim 1 is characterized in that: The segmentation result information of the pixel points where the three 3D hole features are located is merged as follows: The three three-dimensional images are merged, and each pixel in the merged three-dimensional image contains different hole feature information in three directions. The different hole feature information in these three directions is taken as a union, that is, as long as there is a hole feature in the three directions of the pixel, the pixel will be marked as the corresponding hole feature.

8. A three-dimensional digital core pore feature intelligent stereo segmentation method computing system, characterized in that: include: Slice module: used to obtain three-dimensional digital core stereoscopic imaging and slice the three-dimensional digital core stereoscopic imaging into multiple two-dimensional image slices; Segmentation module: input multiple two-dimensional image slices into the trained core pore feature instance segmentation model to obtain the pore feature contours and areas in the multiple two-dimensional image slices; Mapping module: maps the hole feature contours and areas on each slice back to three-dimensional space and reconstructs three-dimensional hole segmentation structures; Merging module: Merge the segmentation result information of the pixel points where the three-dimensional pore and fracture features are located to realize the segmentation and extraction of the pore and fracture features of the three-dimensional digital core.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the intelligent stereoscopic segmentation method for pore and fracture features of three-dimensional digital cores as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the intelligent stereoscopic segmentation method for pore and fracture features of three-dimensional digital cores as claimed in any one of claims 1 to 7 are implemented.

Images