Rock debris three-dimensional CT image segmentation method and device based on graph cut algorithm, equipment and storage medium
By constructing a graph model using a graph cut algorithm and performing quality assessment and optimization, the problems of time-consuming, labor-intensive, and error-prone traditional manual thresholding methods are solved, achieving efficient and high-precision automatic segmentation of rock cutting images.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-25
- Publication Date
- 2026-03-27
AI Technical Summary
Traditional manual thresholding methods are time-consuming, labor-intensive, and highly subjective in cuttings image segmentation, making it difficult to handle complex and variable cuttings images and resulting in large segmentation errors.
A graph-cut algorithm-based segmentation method for 3D CT images of rock cuttings is adopted. Through graph model construction, quality assessment and optimization processing, efficient and high-precision segmentation of rock cuttings images is automatically achieved.
It reduces manual intervention, increases the degree of automation in operation, shortens the segmentation time, and improves the accuracy and efficiency of rock cuttings image segmentation, making it suitable for processing a large number of three-dimensional rock cuttings images.
Smart Images
Figure CN121746404A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of image processing, and is a rock debris three-dimensional CT image segmentation method, device and equipment based on a graph cut algorithm and a storage medium. BACKGROUND
[0002] In the field of geology, the detection of rock debris samples is an important link for understanding the pore structure and physical properties of rocks. In the detection of rock debris samples, a manual threshold segmentation method is often used to segment rock debris images to obtain pore structures. At present, the traditional manual threshold segmentation method is time-consuming, laborious and subjective, and the segmentation result is prone to large errors, and the method is difficult to handle a large number of and complex and variable rock debris images.
[0003] Therefore, how to reduce the time consumption of rock debris image segmentation and improve the accuracy of rock debris image segmentation has become a technical problem to be solved. SUMMARY
[0004] The present application provides a rock debris three-dimensional CT image segmentation method based on a graph cut algorithm, which overcomes the shortcomings of the prior art and effectively solves the problems of poor accuracy and slow speed of the prior art rock debris image segmentation method.
[0005] One of the technical solutions of the present application is achieved by the following measures: a rock debris three-dimensional CT image segmentation method based on a graph cut algorithm, comprising: obtaining a target rock debris three-dimensional image; performing graph model construction on the target rock debris three-dimensional image to obtain a target graph model, wherein the target graph model comprises graph nodes and adjacent edges, the graph nodes correspond to voxels in the target rock debris three-dimensional image, and the adjacent edges connect adjacent graph nodes; performing image segmentation processing on the target graph model to obtain an original image segmentation result; performing quality evaluation on the original image segmentation result according to the graph nodes and the adjacent edges to obtain segmentation quality evaluation data; performing optimization processing on the original image segmentation result according to the segmentation quality evaluation data to obtain a target image segmentation result.
[0006] The following is a further optimization or / and improvement of the above-mentioned one of the technical solutions of the present application: The above-mentioned quality evaluation on the original image segmentation result according to the graph nodes and the adjacent edges to obtain segmentation quality evaluation data comprises: determining a cost value according to the voxel gray value of the graph node and a preset gray threshold value, wherein the cost value is used to represent the possibility that the original image segmentation result belongs to a pore or a rock skeleton; determine a voxel gray value difference according to the graph node, the adjacent edge and the voxel gray value of the graph node; evaluate similarity between different graph nodes according to the voxel gray value difference, to obtain a similarity evaluation value, wherein the similarity evaluation value represents a category to which the original image segmentation result belongs; construct the segmentation quality evaluation data according to the cost value and the similarity evaluation value.
[0007] The method for determining the cost value according to the voxel gray value of the graph node and the preset gray threshold value comprises: If the voxel gray value is greater than the gray threshold value, the cost value is determined as a first cost value; If the voxel gray value is less than or equal to the gray threshold value, the cost value is determined as a second cost value, and the first cost value is greater than the second cost value.
[0008] The method for optimizing the original image segmentation result according to the segmentation quality evaluation data to obtain a target image segmentation result comprises: performing preliminary optimization on the original image segmentation result according to the segmentation quality evaluation data to obtain an optimized image segmentation result; updating the segmentation quality evaluation data according to the optimized image segmentation result and a preset image segmentation algorithm to obtain updated quality evaluation data; until the updated quality evaluation data is minimized, taking the optimized image segmentation result corresponding to the minimized updated quality evaluation data as the target image segmentation result.
[0009] The method for optimizing the original image segmentation result according to the segmentation quality evaluation data to obtain a target image segmentation result comprises: performing step-by-step iteration on the original image segmentation result according to a preset iteration threshold value and the segmentation quality evaluation data to obtain a candidate image segmentation result; until the iteration number reaches the iteration threshold value, taking the candidate image segmentation result of the last iteration as the target image segmentation result.
[0010] The method for performing image segmentation processing on the target graph model to obtain an original image segmentation result comprises: obtaining a gray value difference between the gray value of the graph node and a preset gray threshold value to obtain a gray difference value; performing label allocation on the graph node according to the gray difference value to obtain a category label, wherein the category label is a pore label or a skeleton label; performing region selection on the target graph model according to the pore label to obtain an original pore region; According to the skeleton label, the target graph model is regionally selected to obtain an original skeleton region; According to the original pore region and the original skeleton region, the original image segmentation result is constructed.
[0011] The above obtaining the target rock debris three-dimensional image comprises: Obtaining an original three-dimensional image; The original three-dimensional image is subjected to noise removal to obtain a denoised three-dimensional image; The denoised three-dimensional image is subjected to contrast enhancement to obtain a target three-dimensional image.
[0012] The technical scheme two of the present application is realized by the following measures: a rock debris three-dimensional CT image segmentation device based on a graph cut algorithm, comprising: An image acquisition module acquires a target rock debris three-dimensional image; A model construction module constructs a graph model of the target rock debris three-dimensional image to obtain a target graph model, wherein the target graph model comprises graph nodes and adjacent edges, the graph nodes correspond to voxels in the target rock debris three-dimensional image, and the adjacent edges connect adjacent graph nodes; An original segmentation module performs image segmentation processing on the target graph model to obtain an original image segmentation result; A quality evaluation module evaluates the quality of the original image segmentation result according to the graph nodes and the adjacent edges to obtain segmentation quality evaluation data; An optimization processing module optimizes the original image segmentation result according to the segmentation quality evaluation data to obtain a target image segmentation result.
[0013] The technical scheme three of the present application is realized by the following measures: a storage medium, the storage medium stores a computer program readable by a computer, and the computer program is set to execute the rock debris three-dimensional CT image segmentation method based on the graph cut algorithm when running.
[0014] The technical scheme four of the present application is realized by the following measures: an electronic device comprising a processor and a memory, the memory stores a computer program, the computer program is loaded and executed by the processor to realize the rock debris three-dimensional CT image segmentation method based on the graph cut algorithm.
[0015] The present application provides an image automatic segmentation method based on a graph cut algorithm, realizes high-precision and high-efficiency automatic segmentation of rock debris micron CT three-dimensional images, and provides accurate and reliable data basis for subsequent porosity, pore structure analysis and permeability simulation. BRIEF DESCRIPTION OF DRAWINGS
[0016] ATTACHFigure 1 This is a flowchart of the rock debris image segmentation method according to Embodiment 1 of the present invention.
[0017] Appendix Figure 2 This is a flowchart of step 104 in Embodiment 5 of the present invention.
[0018] Appendix Figure 3 This is a schematic diagram of the structure of the three-dimensional CT image segmentation device for rock cuttings according to the present invention.
[0019] Appendix Figure 4 This is a schematic diagram of the hardware structure of the electronic device of the present invention. Detailed Implementation
[0020] The present invention is not limited to the following embodiments, and the specific implementation can be determined according to the technical solution of the present invention and the actual situation.
[0021] The following is an explanation of some of the terms used in this invention: Three-dimensional images of rock cuttings: These are three-dimensional visualizations of rock cutting samples obtained through various imaging techniques (such as CT scans and 3D reconstruction). These images reveal the internal structural features of the rock cuttings, such as pores and cracks.
[0022] Rock skeleton: refers to the structure of the non-porous part of a rock, which is mainly composed of mineral grains, crystals and other solid components.
[0023] Image segmentation is a technique that divides an image into several specific regions with different properties and extracts the target region.
[0024] Voxels are the basic units that represent data in three-dimensional space. Voxels are widely used in fields such as image processing and geology.
[0025] The present invention will be further described below with reference to embodiments: Example 1: As Figure 1 As shown, this embodiment of the invention discloses a method for segmenting three-dimensional CT images of rock cuttings based on a graph cut algorithm, including: Step 101: Obtain a 3D image of the target rock cuttings; Step 102: Construct a graph model for the target rock debris 3D image to obtain the target graph model. The target graph model includes graph nodes and adjacent edges. The graph nodes correspond to voxels in the target rock debris 3D image, and the adjacent edges connect adjacent graph nodes. Step 103: Perform image segmentation on the target graph model to obtain the original image segmentation result; Step 104: Evaluate the quality of the original image segmentation results based on graph nodes and adjacent edges to obtain segmentation quality evaluation data; Step 105, according to the segmentation quality evaluation data, the original image segmentation result is optimized to obtain the target image segmentation result.
[0026] Figure 1 is an optional flowchart of the cutting image segmentation method provided by the present application, Figure 1 The method in the method can include but is not limited to steps 101-105.
[0027] The beneficial effects of the present application include but are not limited to: according to the target cutting three-dimensional image, the target graph model is constructed, and the image segmentation processing is carried out on the target graph model, and the original image segmentation result is obtained; then, according to the graph node and the adjacent edge of the target graph model, the quality of the original image segmentation result is evaluated, and the segmentation quality evaluation data is obtained; according to the segmentation quality evaluation data, the original image segmentation result is optimized to obtain the target image segmentation result. It can be seen that, after the original image segmentation result is evaluated to obtain the segmentation quality evaluation data, the original image segmentation result is optimized according to the segmentation quality evaluation data, and the target image segmentation result is obtained, which can automatically realize the segmentation of the cutting image, reduce the manual participation, improve the automation degree of operation, and reduce the time consumption of the cutting image segmentation, shorten the segmentation time to make the cutting image segmentation more efficient; improve the accuracy of the cutting image segmentation, and be suitable for the processing scene of segmenting a large number of cutting three-dimensional images.
[0028] Embodiment 2: As an optimization of the above-mentioned embodiment, in step 101, the target cutting three-dimensional image is obtained, including: obtaining an original three-dimensional image; removing noise from the original three-dimensional image to obtain a denoised three-dimensional image; enhancing the contrast of the denoised three-dimensional image to obtain a target three-dimensional image.
[0029] The advantage of the present embodiment is that after obtaining the original three-dimensional image, the original three-dimensional image is preprocessed, including noise removal and contrast enhancement, so as to obtain a target three-dimensional image more suitable for subsequent processing, thereby improving the accuracy of image segmentation.
[0030] Specifically, the target cutting sample can be CT (Computed Tomography, Computer Tomography) scanned to obtain the original cutting three-dimensional image. Specifically, the target cutting sample can be CT scanned by a micron CT scanner. It should be noted that the micron CT scanner is an analytical instrument mainly used for high-precision scanning of three-dimensional pore structure of samples such as cores or cuttings. The target cutting three-dimensional image can also be obtained by other means, which is not limited thereto.
[0031] It should be noted that the target rock debris three-dimensional image has a plurality of voxels. For example, the target rock debris three-dimensional image can be a cube with a side length of 800 grids, and the target rock debris three-dimensional image has a total of 800*800*800=512 million voxels. The side length of each voxel is related to the resolution of the image, and the unit of the side length can be microns.
[0032] It should be noted that each voxel has a gray value. Specifically, it can be a 16-bit gray value, and the gray value can range from 0 to 65535.
[0033] Embodiment 3: As an optimization of the above embodiment, in step 102, the graph node corresponds to the voxel in the target rock debris three-dimensional image, and the gray value of the graph node corresponds to the gray value of the voxel. The adjacent edge represents the spatial relationship between a plurality of voxels, for example, the adjacent edge is used to connect two adjacent graph nodes.
[0034] It should be noted that adjacent voxels can refer to each voxel adjacent to six different voxels above, below, left, right, and front. Adjacent graph nodes also correspond to adjacent voxels. Specifically, if the first voxel and the second voxel in the target rock debris three-dimensional image are adjacent, then the graph node corresponding to the first voxel and the graph node corresponding to the second voxel are also adjacent.
[0035] Embodiment 4: As an optimization of the above embodiment, in step 103, the original image segmentation result can represent the original pore region and the original skeleton region after the target graph model is segmented, specifically, the graph node in the target graph model is marked as a pore or a skeleton.
[0036] Embodiment 5: As shown in Figure 2 As an optimization of the above embodiment, step 104 includes but is not limited to steps 201 to 204: Step 201, determining a cost value according to the voxel gray value of the graph node and the preset gray threshold value, wherein the cost value is used to represent the possibility that the original image segmentation result belongs to the pore or the rock skeleton; Step 202, determining a voxel gray difference value according to the graph node, the adjacent edge, and the voxel gray value of the graph node; Step 203, performing similarity evaluation between different graph nodes according to the voxel gray difference value to obtain a similarity evaluation value, wherein the similarity evaluation value represents the category to which the original image segmentation result belongs; Step 204, constructing segmentation quality evaluation data according to the cost value and the similarity evaluation value.
[0037] The embodiment has the advantages that the original image segmentation result is evaluated in quality according to the graph nodes and the adjacent edges, segmentation quality evaluation data is obtained, the quality of the image segmentation result can be represented by the segmentation quality evaluation data, and the accuracy of the image segmentation is improved.
[0038] In step 202, the adjacent graph nodes of the graph node are determined according to the graph nodes connected at the two ends of the adjacent edge, and the voxel gray value difference is calculated by the graph node and the adjacent graph nodes thereof.
[0039] For example, a graph node has 6 adjacent graph nodes, in which case, the gray value difference between the graph node and each adjacent graph node is calculated to obtain 6 initial gray value differences; and the 6 initial gray value differences are processed by weighted summation to determine the voxel gray value difference.
[0040] In step 203, the similarity evaluation value is a measure of the dissimilarity between the graph node and the adjacent graph nodes thereof. The greater the voxel gray value difference, the greater the similarity evaluation value, indicating that the gray value of the graph node and the gray value of the adjacent graph nodes are more different, and the graph node and the adjacent graph nodes thereof may belong to different categories.
[0041] In step 204, the cost value and the similarity evaluation value are summed to determine the segmentation quality evaluation data.
[0042] In step 201, the following steps are included but not limited to: If the voxel gray value is greater than the gray threshold, the cost value is determined as a first cost value; If the voxel gray value is less than or equal to the gray threshold, the cost value is determined as a second cost value, and the first cost value is greater than the second cost value.
[0043] The embodiment has the advantages that the cost value is determined according to the voxel gray value of the graph node and the gray threshold, and the cost value of the graph node with a smaller voxel gray value is set as a smaller first cost value, so that the segmentation quality evaluation data can be minimized thereafter to determine the target image segmentation result.
[0044] It should be noted that the graph node can be set with a pore label or a skeleton label according to the size relationship between the voxel gray value of the graph node and the preset gray threshold, and the cost value is set according to the pore label or the skeleton label. For example, if the voxel gray value of the graph node is less than the gray threshold, the graph node is set with a pore label, and the graph node is set with a smaller cost value; if the voxel gray value of the graph node is greater than or equal to the gray threshold, the graph node is set with a skeleton label, and the graph node is set with a larger cost value.
[0045] In the embodiment 10, as an optimization of the above-mentioned embodiments, the gray scale threshold includes a pore gray scale threshold and a skeleton gray scale threshold. The pore gray scale correlation degree is obtained by calculating the difference between the voxel gray scale value and the pore gray scale threshold. The skeleton gray scale correlation degree is obtained by calculating the difference between the voxel gray scale value and the skeleton gray scale threshold. If the pore gray scale correlation degree is less than the skeleton gray scale correlation degree, the pore label is set to the graph node, and the first cost value is determined as the cost value. If the pore gray scale correlation degree is greater than or equal to the skeleton gray scale correlation degree, the skeleton label is set to the graph node, and the second cost value is determined as the cost value.
[0046] In the embodiment 11, as an optimization of the above-mentioned embodiments, the target graph model further includes a first terminal node, a second terminal node and a category edge. The gray scale value of the first terminal node is the pore threshold, and the gray scale value of the second terminal node is the skeleton threshold. The category edge is used to connect the graph node with the first terminal node or connect the graph node with the second terminal node. The graph node connected with the first terminal node is equivalent to the graph node with the pore label, that is, marked as the pore. The graph node connected with the second terminal node is equivalent to the graph node with the skeleton label, that is, marked as the skeleton.
[0047] In the embodiment 12, as an optimization of the above-mentioned embodiments, the step 105 includes but is not limited to the following steps. performing preliminary optimization processing on the original image segmentation result according to the segmentation quality evaluation data to obtain an optimized image segmentation result; updating the segmentation quality evaluation data according to the optimized image segmentation result and the preset image segmentation algorithm to obtain updated quality evaluation data; until the updated quality evaluation data is minimized, the optimized image segmentation result corresponding to the minimized updated quality evaluation data is taken as the target image segmentation result.
[0048] The embodiment has the advantages that the segmentation quality evaluation data is updated according to the optimized image segmentation result and the preset image segmentation algorithm to obtain the updated quality evaluation data, and the optimized image segmentation result corresponding to the minimized updated quality evaluation data is taken as the target image segmentation result, so that the image segmentation is realized through the image segmentation algorithm, and the efficiency and accuracy of the image segmentation are improved.
[0049] It should be noted that the image segmentation algorithm can be a machine learning algorithm or a deep learning algorithm, and the embodiment of the present application does not limit the image segmentation algorithm.
[0050] In the embodiment 13, as an optimization of the above-mentioned embodiments, the step 105 of performing optimization processing on the original image segmentation result according to the segmentation quality evaluation data to obtain the target image segmentation result includes the following steps. Based on the preset iteration threshold and segmentation quality evaluation data, the original image segmentation results are iterated step by step to obtain candidate image segmentation results; The process continues until the iteration count reaches the iteration threshold, at which point the candidate image segmentation result of the last iteration is used as the target image segmentation result.
[0051] The advantage of this embodiment is that it iterates the original image segmentation result step by step according to the preset iteration threshold and segmentation quality evaluation data, and uses the candidate image segmentation result of the last iteration as the target image segmentation result, thereby realizing image segmentation through the image segmentation algorithm and improving the efficiency and accuracy of image segmentation.
[0052] It should be noted that the segmentation quality assessment data is updated after each iteration, and the updated segmentation quality assessment data is used as the segmentation quality assessment data for the next iteration.
[0053] Example 14: As an optimization of the above embodiment, step 103 includes, but is not limited to: Obtain the difference between the grayscale value of a graph node and the preset grayscale threshold to get the grayscale difference. The graph nodes are labeled according to the grayscale difference to obtain category labels, where the category labels are either pore labels or skeleton labels. Based on the pore labels, the target map model is selected to obtain the original pore region; Based on the skeleton labels, regions are selected from the target graph model to obtain the original skeleton regions; The original image segmentation result is constructed based on the original pore region and the original skeleton region.
[0054] The advantage of this embodiment is that it performs preliminary image segmentation on the three-dimensional image of the target rock debris to obtain the original image segmentation result, which helps to improve the processing efficiency of image segmentation.
[0055] The automatic segmentation method for three-dimensional images of rock cuttings from micron-CT scans in this invention significantly improves analysis efficiency and reduces manual operation time. The application of the graph-cut algorithm improves segmentation accuracy, overcomes the limitations of traditional thresholding methods, and ensures the objectivity and accuracy of the results. Furthermore, this method is applicable to large-scale micron-CT scan data, providing a powerful analytical tool for geology, oil and gas exploration, and other fields, and promoting in-depth research in these areas.
[0056] Example 15: As Figure 3 As shown, this embodiment of the invention discloses a rock cuttings image segmentation device, which can implement the above-mentioned rock cuttings image segmentation method. The device includes: Image acquisition module 301 acquires a three-dimensional image of the target rock cuttings; The model building module 302 constructs a graph model from the target rock debris 3D image to obtain the target graph model. The target graph model includes graph nodes and adjacent edges. The graph nodes correspond to voxels in the target rock debris 3D image, and the adjacent edges connect adjacent graph nodes. The original segmentation module 303 performs image segmentation processing on the target graph model to obtain the original image segmentation result; The quality assessment module 304 performs quality assessment on the original image segmentation results based on graph nodes and adjacent edges to obtain segmentation quality assessment data. The optimization processing module 305 optimizes the original image segmentation result based on the segmentation quality assessment data to obtain the target image segmentation result.
[0057] The specific implementation of this rock debris image segmentation device is basically the same as the specific implementation of the rock debris image segmentation method described above, and will not be repeated here.
[0058] Example 16: As Figure 3 As shown in the figure, an embodiment of the present invention discloses an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-described rock debris image segmentation method. This electronic device can be any smart terminal, including tablet computers, in-vehicle computers, etc.
[0059] like Figure 4 As shown, Figure 4 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes: The processor 401 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application. The memory 402 can be implemented as a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 402 can store the operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 402 and is called and executed by the processor 401 using the rock cuttings image segmentation method of the embodiments of this application. Input / output interface 403 is used to implement information input and output; The communication interface 404 is configured to realize the communication interaction between the device and other devices, and can realize the communication through a wired manner (for example, a USB, a network cable and the like) or a wireless manner (for example, a mobile network, WIFI, Bluetooth and the like). The bus 405 is configured to transmit information between various components (for example, the processor 401, the memory 402, the input / output interface 403 and the communication interface 404) of the device. The processor 401, the memory 402, the input / output interface 403 and the communication interface 404 are connected to each other through the bus 405.
[0060] The processor described above can be a central processing unit (CPU), a general-purpose processor, a digital signal processor (DSP), an ASIC, an FPGA or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. The processor can realize or execute various exemplary logical blocks, modules and circuits described in combination with the disclosure. The processor can also be a combination of computing functions, such as one or more microprocessor combinations, combinations of DSP and microprocessor, and the like. The memory can include, but is not limited to, a U disk, a read-only memory, a mobile hard disk, a magnetic disk or an optical disk and various computer program storage media.
[0061] The present application can acquire and process related data based on artificial intelligence technology. The artificial intelligence (AI) is to use a digital computer or a machine controlled by a digital computer to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use the knowledge to obtain the best results.
[0062] The basic technology of artificial intelligence generally includes technologies such as sensors, special artificial intelligence chips, cloud computing, distributed storage, big data processing, operation / interaction systems, mechatronics and the like. The software technology of artificial intelligence mainly includes computer vision technology, robot technology, biometric technology, speech processing technology, natural language processing technology and machine learning / deep learning and the like.
[0063] Those skilled in the art will appreciate that embodiments of the application can be readily used as a method, a system or a computer program product. Accordingly, the application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the application can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) embodying computer-readable program code. Embodiments of the application are not described with reference to any particular programming language. It will be appreciated that a variety of programming languages can be used to implement the teachings of the application as described herein.
[0064] The computer program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other processing devices to cause a series of operational steps to be performed on the computer, other programmable apparatus or other processing devices to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions specified in the flowchart block or blocks. Figure 1 The flowchart and / or block diagram in the Figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. In this regard, each flowchart block and / or block in the Figures can represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical functions. It should also be noted that the flowchart blocks and / or blocks in the Figures can represent a phase of a method, a segment of a function, and / or a portion of a computer program product. Figure 1 The computer program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other processing devices to cause a series of operational steps to be performed on the computer, other programmable apparatus or other processing devices to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions specified in the flowchart block or blocks. The computer program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other processing devices to cause a series of operational steps to be performed on the computer, other programmable apparatus or other processing devices to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions specified in the flowchart block or blocks.
[0065] The computer program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other processing devices to cause a series of operational steps to be performed on the computer, other programmable apparatus or other processing devices to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions specified in the flowchart block or blocks. Figure 1 The computer program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other processing devices to cause a series of operational steps to be performed on the computer, other programmable apparatus or other processing devices to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions specified in the flowchart block or blocks. Figure 1 The computer program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other processing devices to cause a series of operational steps to be performed on the computer, other programmable apparatus or other processing devices to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions specified in the flowchart block or blocks. The computer program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other processing devices to cause a series of operational steps to be performed on the computer, other programmable apparatus or other processing devices to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions specified in the flowchart block or blocks.
[0066] The above technical features constitute the best mode of the present application, which has strong adaptability and best implementation effect. Non-essential technical features can be added or removed according to actual needs to meet the needs of different situations.
Claims
1. A method for segmenting a three-dimensional CT image of debris based on a graph cut algorithm, characterized in that include: Acquire a 3D image of the target rock cuttings; A graph model is constructed on the target rock debris 3D image to obtain a target graph model, wherein the target graph model includes: graph nodes and adjacent edges, the graph nodes correspond to voxels in the target rock debris 3D image, and the adjacent edges connect adjacent graph nodes; The target graph model is subjected to image segmentation processing to obtain the original image segmentation result; The original image segmentation result is evaluated based on the graph nodes and the adjacent edges to obtain segmentation quality evaluation data; The original image segmentation result is optimized based on the segmentation quality assessment data to obtain the target image segmentation result. 2.The method of claim 1, wherein The quality of the original image segmentation result is evaluated based on the graph nodes and the adjacent edges to obtain segmentation quality evaluation data, including: The cost value is determined based on the voxel gray value of the graph node and a preset gray value threshold, wherein the cost value is used to characterize the probability that the original image segmentation result belongs to pores or rock skeleton. The voxel grayscale difference is determined based on the graph node, the adjacent edge, and the voxel grayscale value of the graph node. The similarity between different graph nodes is evaluated based on the voxel gray-level difference to obtain a similarity evaluation value, wherein the similarity evaluation value represents the category to which the original image segmentation result belongs; The segmentation quality assessment data is constructed based on the cost value and the similarity assessment value. 3.The method of claim 2, wherein The cost value is determined based on the voxel grayscale value of the graph node and a preset grayscale threshold, including: If the voxel gray value is greater than the gray value threshold, the cost value is determined to be the first generation value; If the voxel gray value is less than or equal to the gray value threshold, the cost value is determined to be the second-generation value, and the first-generation value is greater than the second-generation value.
4. The rock debris 3D CT image segmentation method based on the graph cut algorithm according to any one of claims 1 to 3, characterized in that The original image segmentation result is optimized based on the segmentation quality assessment data to obtain the target image segmentation result, including: The original image segmentation result is initially optimized based on the segmentation quality assessment data to obtain an optimized image segmentation result. The segmentation quality assessment data is updated based on the optimized image segmentation results and the preset image segmentation algorithm to obtain updated quality assessment data. The process continues until the updated quality assessment data is minimized, at which point the optimized image segmentation result corresponding to the minimized updated quality assessment data is taken as the target image segmentation result.
5. The rock debris 3D CT image segmentation method based on the graph cut algorithm according to any one of claims 1 to 3, characterized in that The original image segmentation result is optimized based on the segmentation quality assessment data to obtain the target image segmentation result, including: The original image segmentation result is iterated step by step according to the preset iteration threshold and the segmentation quality evaluation data to obtain candidate image segmentation results; The process continues until the iteration count reaches the iteration threshold, at which point the candidate image segmentation result of the last iteration is taken as the target image segmentation result.
6. The rock debris 3D CT image segmentation method based on the graph cut algorithm according to any one of claims 1 to 5, characterized in that The target graph model is subjected to image segmentation processing to obtain the original image segmentation result, including: The difference between the grayscale value of the graph node and the preset grayscale threshold is obtained to obtain the grayscale difference value; According to the gray difference value, the graph node is labeled and assigned to obtain a category label, wherein the category label is a pore label or a skeleton label; According to the pore label, a region selection is performed on the target graph model to obtain an original pore region; According to the skeleton label, a region selection is performed on the target graph model to obtain an original skeleton region; According to the original pore region and the original skeleton region, the original image segmentation result is constructed.
7. The rock debris 3D CT image segmentation method based on the graph cut algorithm according to any one of claims 1 to 6, characterized in that A target rock debris three-dimensional image is acquired, including: An original three-dimensional image is acquired; Noise removal is performed on the original three-dimensional image to obtain a denoised three-dimensional image; Contrast enhancement is performed on the denoised three-dimensional image to obtain a target three-dimensional image.
8. A device for segmenting a three-dimensional CT image of debris based on a graph cut algorithm, which uses the method for segmenting a three-dimensional CT image of debris based on a graph cut algorithm according to any one of claims 1 to 7, characterized by It includes: An image acquisition module acquires a target rock debris three-dimensional image; A model construction module constructs a graph model from the target rock debris three-dimensional image to obtain a target graph model, wherein the target graph model includes graph nodes and adjacent edges, the graph nodes correspond to voxels in the target rock debris three-dimensional image, and the adjacent edges connect adjacent graph nodes; An original segmentation module performs image segmentation processing on the target graph model to obtain an original image segmentation result; A quality evaluation module evaluates the quality of the original image segmentation result according to the graph nodes and the adjacent edges to obtain segmentation quality evaluation data; An optimization processing module optimizes the original image segmentation result according to the segmentation quality evaluation data to obtain a target image segmentation result.
9. A storage medium characterized by The storage medium has a computer program readable by a computer stored thereon, and the computer program is configured to execute the rock debris three-dimensional CT image segmentation method based on the graph cut algorithm as claimed in any one of claims 1 to 7 when running.
10. An electronic device, characterized in that... It includes a processor and a memory, and the memory has a computer program stored therein, and the computer program is loaded and executed by the processor to realize the rock debris three-dimensional CT image segmentation method based on the graph cut algorithm as claimed in any one of claims 1 to 7.