Phase segmentation method and electronic device for composite cement system based on machine vision

Abstract

The present application provides a method and electronic device for phase segmentation of a composite cement system based on machine vision. The method includes: obtaining a BSE image of a composite cement system as an image data set, wherein the composite cement system contains a variety of cementitious materials including cement, and each BSE image includes a training part and a segmentation part, wherein the training part is marked with ROI, and the segmentation part is not marked with ROI, and the ROI mark is used to mark the classification labels of one or more phases; extracting different features of the BSE image by different feature extraction methods to obtain feature stacks of different features; training and optimizing a first random forest algorithm based on the classification labels and feature stacks of the training part; performing visual phase segmentation on the segmentation part based on the first random forest algorithm and the feature stack of the segmentation part, and distinguishing one or more phases in the segmentation part from the result of the visual phase segmentation.

Description

Technical Field

[0001] The present application relates to the field of machine vision technology, and in particular to a method and electronic device for phase segmentation of a composite cement system based on machine vision. Background Art

[0002] The use of mineral admixtures to replace cement clinker in cement concrete can reduce the amount of cement used and has good environmental benefits. However, the addition of mineral admixtures makes the hydration process of the composite cement system random and complicated, making the hydration mechanism difficult to clarify. In the study of the mechanism and properties of cement-based materials, images can often intuitively provide rich and valuable phase information, such as back-scattered electron (BSE) images of cement-based materials. Grayscale segmentation of cement-based material images to distinguish different phases in cement-based materials has been widely studied. However, for composite cement systems with complex phase types, the grayscales of different phases will overlap, so grayscale segmentation may be difficult to accurately distinguish the phases. Summary of the invention

[0003] The embodiments of the present application provide a method and an electronic device for phase segmentation of a composite cement system based on machine vision.

[0004] In a first aspect, an embodiment of the present application provides a method for phase segmentation of a composite cement system based on machine vision, the method comprising:

[0005] Acquire BSE images of a composite cement system as an image data set. The composite cement system contains a variety of cementitious materials including cement. Each BSE image includes a training part and a segmentation part. The training part is marked with ROI, and the segmentation part is not marked with ROI. The ROI marking is used to mark the classification labels of one or more phases.

[0006] Different features of the BSE image are extracted by different feature extraction methods to obtain feature stacks with different features;

[0007] Training and optimizing a first random forest algorithm based on the classification labels and feature stacks of the training part, the first random forest algorithm is used to perform visual phase segmentation of the composite cement system;

[0008] The segmented part is subjected to visual phase segmentation based on the first random forest algorithm and the feature stack of the segmented part, and one or more phases in the segmented part are distinguished in the result of the visual phase segmentation.

[0009] Optionally, the method further includes:

[0010] Convert the result of visual phase segmentation into a binary image;

[0011] Calculating the volume fraction of one or more phases based on the area fraction of one or more phases in the binary image;

[0012] Based on the volume fraction, the porosity in the composite cement system and the hydration degree of various cementitious materials were calculated.

[0013] Optionally, the first random forest algorithm is also used to perform visual phase segmentation on phases of specified sizes and / or phases of specified classification in the composite cement system.

[0014] Optionally, different feature extraction methods include Gaussian filtering, Sober filtering and pixel averaging, and different features include texture features, color features, edge features, and contrast features of different phases in the BSE image.

[0015] In a second aspect, an embodiment of the present application provides an electronic device for performing phase segmentation of a composite cement system based on machine vision, the electronic device comprising:

[0016] An acquisition module is used to acquire BSE images of a composite cement system as an image data set. The composite cement system contains a variety of cementitious materials including cement. Each BSE image includes a training part and a segmentation part. The training part is marked with ROI, and the segmentation part is not marked with ROI. The ROI mark is used to mark the classification labels of one or more phases.

[0017] An extraction module is used to extract different features of the BSE image through different feature extraction methods to obtain feature stacks of different features;

[0018] A training and optimization module, used for training and optimizing a first random forest algorithm based on the classification labels and feature stack of the training part, wherein the first random forest algorithm is used for visualizing physical phase segmentation of the composite cement system;

[0019] The segmentation module is used to perform visual phase segmentation on the segmented part based on the first random forest algorithm and the feature stack of the segmented part, and distinguish one or more phases in the segmented part from the result of the visual phase segmentation.

[0020] Optionally, the electronic device further includes:

[0021] A conversion module, used to convert the result of visual phase segmentation into a binary image;

[0022] A first calculation module, used to calculate the volume fraction of one or more phases based on the area fraction of one or more phases in the binary image;

[0023] The second calculation module is used to calculate the porosity in the composite cement system and the hydration degree of various cementitious materials based on the volume fraction.

[0024] Optionally, the first random forest algorithm is also used to perform visual phase segmentation on phases of specified sizes and / or phases of specified classification in the composite cement system.

[0025] Optionally, different feature extraction methods include Gaussian filtering, Sober filtering and pixel averaging, and different features include texture features, color features, edge features, and contrast features of different phases in the BSE image.

[0026] In a third aspect, an embodiment of the present application provides an electronic device comprising a memory, at least one processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, a method as described in any one of the above-mentioned first aspects is implemented.

[0027] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method as described in any one of the above-mentioned first aspects is implemented.

[0028] In a fifth aspect, an embodiment of the present application provides a computer program product. When the computer program product is run on an electronic device, the electronic device executes any one of the methods described in the first aspect.

[0029] Based on the present application, each BSE image of the composite cement system can be divided into a training part and a segmentation part, and the classification labels and feature stacks of the training part can be used to train and optimize the first random forest algorithm, so that the first random forest algorithm learns the correspondence between the classification labels and feature stacks in each BSE image, thereby helping the first random forest algorithm to accurately perform visual phase segmentation on the segmentation part according to the feature stack of the segmentation part. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for describing the embodiments are briefly introduced below.

[0031] Figure 1 It is a flow chart of a method for phase segmentation of a composite cement system based on machine vision provided in an embodiment of the present application.

[0032] Figure 2 It is a schematic diagram of the process of phase separation of cement paste provided in an embodiment of the present application.

[0033] Figure 3 It is a schematic diagram of the morphology of different cementitious particles in the BSE image of the cement paste provided in the examples of the present application.

[0034] Figure 4It is a schematic diagram of ROI marking of the training part in the BSE image of the cement paste provided in the embodiment of the present application.

[0035] Figure 5 It is a characteristic diagram of different phases in the BSE image of the cement paste provided in the embodiment of the present application.

[0036] Figure 6 This is an example diagram of the phase segmentation results of the BSE image of the cement paste provided in the embodiment of the present application.

[0037] Figure 7 It is a schematic diagram of the process of processing the phase segmentation results of the BSE image of cement paste provided in the embodiment of the present application.

[0038] Figure 8 It is a schematic diagram of the structure of an electronic device provided in one embodiment of the present application.

[0039] Fig. 9 It is a structural schematic diagram of an electronic device provided in another embodiment of the present application. DETAILED DESCRIPTION

[0040] The technical solution of the embodiment of the present invention will be described below in conjunction with the accompanying drawings in the embodiment of the present invention.

[0041] The use of mineral admixtures to replace cement clinker in cement concrete can reduce the amount of cement used and has good environmental benefits. However, the addition of mineral admixtures makes the hydration process of the composite cement system random and complicated, making the hydration mechanism difficult to clarify. In the study of the mechanism and properties of cement-based materials, images can often intuitively provide rich and valuable phase information, such as BSE images of cement-based materials. Grayscale segmentation of cement-based material images to distinguish different phases in cement-based materials has been widely studied. However, for composite cement systems with complex phase types, the grayscales of different phases will overlap, so grayscale segmentation may be difficult to accurately distinguish the phases.

[0042] Based on this, Figure 1 As shown in an embodiment of the present application, a method for phase segmentation of a composite cement system based on machine vision is provided, and the method may include steps S110-S140.

[0043] In step S110, a BSE image of a composite cement system is obtained as an image data set. The composite cement system may contain a variety of cementitious materials including cement, such as slag, fly ash, aggregate (such as sand), cement, etc. After the composite cement system is formed, different cementitious materials will undergo hydration reactions to produce hydration products. Therefore, the physical phases in the composite cement system may include a variety of cementitious materials, hydration products, and pores.

[0044] Exemplarily, the BSE image may be a BSE image obtained by preparing a cement test block hardened slurry from a composite cement system, and then crushing, grinding, and pouring epoxy resin. The image data set may include BSE images of any age of the composite cement system, such as BSE images of all ages. In addition, the BSE images in the image data set may be pre-processed such as cropping.

[0045] Each BSE image in the image data set may include a training part and a segmentation part. The training part is marked with a region of interest (ROI), and the segmentation part is not marked with an ROI. The ROI mark is used to mark the classification labels of one or more phases. For example, the classification labels of different phases in the BSE image can be determined based on their visual features (such as edge features, texture features, color features, etc.). The naming of the classification labels is not limited. For example, Arabic numerals can be used to divide them into class 0, class 1, etc. The ROI marking method is also not limited. For example, it can be marked in the BSE image by box selection, point selection, and line selection. In addition, the area ratio of the training part and the segmentation part in the BSE image is not limited.

[0046] In step S120, different features of the BSE image are extracted by different feature extraction methods to obtain feature stacks of different features. The feature stack can be formed by stacking feature maps extracted from the BSE image by different feature extraction methods. Optionally, different feature extraction methods may include Gaussian filtering (i.e., Gaussian smoothing (gaussion blur)), Sobel filtering and pixel averaging, and different features may include texture features, color features, edge features, and contrast features of different phases in the BSE image. Among them, gaussion blur uses a Gaussian function for convolution, so as to achieve a smoothing effect. Sobel filtering can use a Sobel filter to reduce the amount of data, which can not only eliminate irrelevant information, but also retain important structural attributes of the image, and pixel averaging can achieve a denoising effect.

[0047] In step S130, the first random forest algorithm is trained and optimized based on the classification labels and feature stack of the training part. The first random forest algorithm can be used to perform visual phase segmentation on the composite cement system. The advantage of the machine learning algorithm is that it can quickly realize the classification of image phases. This operation is repeatable and greatly improves the efficiency of image processing. When training the first random forest algorithm based on the classification labels and feature stack of the training part, the features and labels can be converted into tensors that can be read by the first random forest algorithm and input into it for training, so that the first random forest algorithm learns the correspondence between features and classification labels. As an example, the first random forest algorithm can be optimized based on the out-of-bag error to improve the segmentation performance of the first random forest algorithm.

[0048] In step S140, the segmented part is visually segmented based on the first random forest algorithm and the feature stack of the segmented part, and one or more phases in the segmented part are distinguished in the result of the visual phase segmentation. The first random forest algorithm can segment multiple phases of the segmented part at one time, such as particles of different cementitious materials, particles of hydration products, or pores. The first random forest algorithm can also segment one phase of the segmented part at one time, which can improve the segmentation accuracy of the first random forest algorithm. After multiple segmentations, multiple phases of the segmented part can be segmented separately. The visualized segmentation result can refer to the distinction and display of different segmented phases in the BSE image, for example, using different colors to represent different phases. Furthermore, the segmentation of phases in the visualized segmentation result can be at the pixel level.

[0049] Optionally, the first random forest algorithm can also be used to perform visual phase segmentation on phases of specified sizes and / or specified classifications in a composite cement system. For example, the first random forest algorithm can be specified to segment phases of a specific classification, such as cement particles. As another example, the first random forest algorithm can also be specified to segment phases of a specific size, such as phases with a particle size of 1 nm. As another example, the first random forest algorithm can also be specified to segment specific phases of a specific size, such as cement particles with a particle size of 1 nm. Based on this, the segmentation results of the first random forest algorithm can be specified according to research needs, which helps to improve research efficiency. For example, specifying the first random forest algorithm to segment cement particles of a specific size can help study the influence of cement grading in a composite cement system.

[0050] Based on the present application, each BSE image of the composite cement system can be divided into a training part and a segmentation part, and the classification labels and feature stacks of the training part can be used to train and optimize the first random forest algorithm, so that the first random forest algorithm learns the correspondence between the classification labels and feature stacks in each BSE image, thereby helping the first random forest algorithm to accurately perform visual phase segmentation on the segmentation part according to the feature stack of the segmentation part.

[0051] Optional, Figure 1 The method shown may further include steps S150 - S170 .

[0052] In step S150, the result of the visual phase segmentation is converted into a binary image. For example, the image predicted by the first random forest algorithm can be converted into a three-primary color RGB image, and after the RGB image is converted into a grayscale image, the image threshold is adjusted to convert it into a binary image.

[0053] In step S160, based on the area fraction of one or more phases in the binary image, the volume fraction of one or more phases is calculated. According to the plane stereology and statistical principles proposed in related research, in a sufficiently small image field of view, the area fraction of a two-dimensional cross section is approximately equal to the volume of a three-dimensional space. Therefore, by calculating the area fraction of one or more phases in the BSE image, the volume of one or more phases can be obtained. Furthermore, the volume fraction of each phase can be calculated based on the volume of multiple phases in the BSE image.

[0054] In step S170, the porosity and the hydration degree of the various cementitious materials in the composite cement system are calculated based on the volume fraction of one or more phases. For example, the porosity in the composite cement system can be calculated based on the volume fraction of the pores. For another example, the hydration degree of the various cementitious materials in the composite cement system can be calculated based on the volume fraction of the unhydrated cementitious materials. As an example, the hydration degree of the various cementitious materials in the composite cement system can be determined based on formula (1).

[0055] (1),

[0056] in, Indicates hydration The first time in the composite cement system The unhydrated volume fraction of the gelling component, Indicates the initial moment of hydration The volume fraction of the cementitious component in the hydrated composite cement system is It indicates the volume change of the composite cement system after the bubbles are introduced into the mixing and the hydration and hardening of the composite cement system. When the mixed composite cement system has good fluidity and is vibrated sufficiently, the bubbles introduced can be ignored, and the volume change under standard curing is not large, so You can choose 1.

[0057] Combine the following Figure 2-Figure 7 Taking a composite cement system (hereinafter referred to as cement paste) including slag, fly ash, aggregate (quartz powder), and cement as an example, the phase segmentation method of the composite cement system based on machine vision in an embodiment of the present application is introduced by way of example.

[0058] Figure 2 It is a schematic diagram of the process of phase separation of cement paste, including steps S210-S260.

[0059] In step S210, an image dataset of cement paste is constructed, including BSE images of cement paste at all ages. The BSE images are obtained after cement paste is prepared as a cement block hardening paste, crushed, polished, poured with epoxy resin, and then photographed. As an example, the morphology of different cementitious particles in the BSE image of cement paste can be as follows: Figure 3As shown, different morphologies have typical texture features, color features, and edge features, so they can be distinguished visually. Figure 3 Among them, A is slag particles, B is cement clinker particles, C is fly ash particles, and D is quartz powder particles.

[0060] In step S220, the training part of each BSE image in the image data set is marked with ROI to obtain the classification label of the marked area. The classification labels of slag, fly ash, aggregate (quartz powder), cement, hydration product, and pore can be class 0, class 1, class 2, class 3, class 4, and class 5, respectively. As an example, the ROI label of the training part in the BSE image can be as follows: Figure 4 shown.

[0061] In step S230, the BSE image is processed by Gaussian filtering, Sobel filtering, and pixel averaging to extract the texture, color, edge features, and contrast features of cement, slag, fly ash, hydration products, and pores. As an example, the feature maps corresponding to cement, slag, fly ash, and quartz powder can be respectively as follows: Figure 5 As shown in A, B, C, and D.

[0062] In step S240, the segmented portion is visually segmented based on the first random forest algorithm and the feature stack of the segmented portion of the BSE image. As an example, the segmentation results of the unhydrated cement, unhydrated slag and unhydrated fly ash in the segmented portion can be as follows: Figure 6 shown.

[0063] In step S250, after the segmentation result image is converted into a grayscale image, the image threshold is adjusted to convert it into a binary image, and the binary image is denoised. As an example, the processing process of the segmentation result image can be as follows: Figure 7 shown.

[0064] In step S260, the porosity of the cement paste and the hydration degree of the various cementitious materials are calculated based on the area fraction of one or more phases in the binary graph. The volume fraction of the phase can be calculated from the area fraction of the phase, and the porosity of the cement paste and the hydration degree of the various cementitious materials can be further calculated.

[0065] based on Figure 2 The method shown can quantitatively analyze the hydration degree of various cementitious materials in composite cement quickly, efficiently, at low cost, and intuitively and clearly.

[0066] The method provided in the embodiment of the present application is described in detail above. The device provided in the embodiment of the present application is described in detail below with reference to the accompanying drawings.

[0067] See also Figure 8 , is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 8 The electronic device 800 shown may include an acquisition module 810 , an extraction module 820 , a training and optimization module 830 , and a segmentation module 840 .

[0068] The acquisition module 810 can be used to acquire BSE images of a composite cement system as an image data set. The composite cement system contains a variety of cementitious materials including cement. Each BSE image includes a training part and a segmentation part. The training part is marked with ROI, and the segmentation part is not marked with ROI. The ROI mark is used to mark the classification labels of one or more phases.

[0069] The extraction module 820 may be used to extract different features from the BSE image using different feature extraction methods to obtain feature stacks of different features;

[0070] The training and optimization module 830 can be used to train and optimize a first random forest algorithm based on the classification labels and feature stack of the training part, and the first random forest algorithm is used to perform visual phase segmentation on the composite cement system;

[0071] The segmentation module 840 may be used to perform visual phase segmentation on the segmented portion based on the first random forest algorithm and the feature stack of the segmented portion, and distinguish one or more phases in the segmented portion from the result of the visual phase segmentation.

[0072] Optionally, the electronic device may further include a conversion module 850 , a first calculation module 860 and a second calculation module 870 .

[0073] The conversion module 850 may be used to convert the result of the visual phase segmentation into a binary image;

[0074] The first calculation module 860 may be used to calculate the volume fraction of one or more phases based on the area fraction of one or more phases in the binary image;

[0075] The second calculation module 870 may be used to calculate the porosity and hydration degree of various cementitious materials in the composite cement system based on the volume fraction.

[0076] Optionally, the first random forest algorithm can also be used to perform visual phase segmentation on phases of specified sizes and / or phases of specified classification in the composite cement system.

[0077] Optionally, different feature extraction methods may include Gaussian filtering, Sober filtering and pixel averaging, and different features may include texture features, color features, edge features, and contrast features of different phases in the BSE image.

[0078] The present application also provides an electronic device. Fig. 9As shown, the electronic device 900 includes: at least one processor 901, a memory 902, and a computer program 903 stored in the memory 902 and executable on the at least one processor 901. When the processor 901 executes the computer program 903, the method provided in the present application is implemented.

[0079] Exemplarily, the computer program 903 may be divided into one or more modules / units, one or more modules / units are stored in the memory 902, and are executed by the processor 901 to complete the present application. One or more modules / units may be a series of computer program instruction segments capable of completing specific functions, and the instruction segments are used to describe the execution process of the computer program in the electronic device 900.

[0080] Those skilled in the art will understand that Fig. 9 It is only an example of the electronic device 900 and does not constitute a limitation of the electronic device. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the electronic device 900 may also include input and output devices, network access devices, buses, etc.

[0081] The processor 901 may 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. A general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc.

[0082] The memory 902 may be an internal storage unit of the electronic device 900, such as a hard disk or memory of the electronic device 900. The memory 902 may also be an external storage device of the electronic device 900, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device 900. Further, the memory 902 may also include both an internal storage unit of the electronic device 900 and an external storage device. The memory 902 is used to store computer programs and other programs and data required by the electronic device 900. The memory 902 may also be used to temporarily store data that has been output or is to be output.

[0083] The electronic device 900 provided in this embodiment can execute the above method embodiment, and its implementation principle and technical effect are similar, which will not be repeated here.

[0084] An embodiment of the present application further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0085] An embodiment of the present application provides a computer program product. When the computer program product runs on an electronic device, the electronic device can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0086] If the integrated unit 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 application implements all or part of the processes in the above method embodiments, which can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above method embodiments can be implemented. Among them, the computer program includes computer program code, which can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device that can carry the computer program code to the camera / electronic device, recording medium, computer memory, ROM (read-only memory), RAM (random access memory), CD-ROM (compact disc read-only memory), magnetic tape, floppy disk and optical data storage device. The computer-readable storage medium mentioned in the present application can be a non-volatile storage medium, in other words, it can be a non-transient storage medium.

[0087] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0088] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0089] In the embodiments provided in the present application, it should be understood that the disclosed devices / equipment and methods can be implemented in other ways. For example, the device / equipment embodiments described above are only schematic, for example, the division of modules or units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0090] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0091] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0092] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0093] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0094] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for phase segmentation of a composite cement system based on machine vision, characterized in that: include: Acquire backscattered electron (BSE) images of a composite cement system as an image data set, wherein the composite cement system contains a variety of cementitious materials including cement, each of the BSE images includes a training part and a segmented part, wherein the training part is marked with a region of interest (ROI), and the segmented part is not marked with the ROI, and the ROI mark is used to mark the classification labels of one or more phases; Extracting different features from the BSE image by different feature extraction methods to obtain feature stacks of the different features, wherein the different features include texture features, color features, edge features, and contrast features of each phase in the BSE image; Training and optimizing a first random forest algorithm based on the classification labels and the feature stack of the training portion, wherein the first random forest algorithm is used to perform visual phase segmentation on the composite cement system; The segmented portion is subjected to visual phase segmentation based on the first random forest algorithm and the feature stack of the segmented portion, and one or more phases in the segmented portion are distinguished in a result of the visual phase segmentation.

2. The method according to claim 1, characterized in that The method further comprises: Converting the result of the visual phase segmentation into a binary image; Calculating the volume fraction of the one or more phases based on the area fraction of the one or more phases in the binary image; Based on the volume fractions, the porosity in the composite cement system and the degree of hydration of the plurality of cementitious materials are calculated.

3. The method according to claim 2, characterized in that: The first random forest algorithm is also used to perform visual phase segmentation on phases of specified sizes and / or phases of specified classification in the composite cement system.

4. The method according to claim 3, characterized in that The different feature extraction methods include Gaussian filtering, Sobel filtering and pixel averaging.

5. An electronic device, characterized in that: The electronic device is used for performing phase segmentation of a composite cement system based on machine vision, and the electronic device comprises: An acquisition module, used for acquiring backscattered electron (BSE) images of a composite cement system as an image data set, wherein the composite cement system contains a variety of cementitious materials including cement, and each of the BSE images includes a training part and a segmentation part, wherein the training part is marked with a region of interest (ROI), and the segmentation part is not marked with the ROI, and the ROI mark is used to mark the classification labels of one or more phases; An extraction module, used to extract different features from the BSE image by using different feature extraction methods to obtain feature stacks of the different features, wherein the different features include texture features, color features, edge features, and contrast features of each phase in the BSE image; A training and optimization module, used for training and optimizing a first random forest algorithm based on the classification labels and the feature stack of the training part, wherein the first random forest algorithm is used for visualizing physical phase segmentation of the composite cement system; A segmentation module is used to perform visual phase segmentation on the segmented part based on the first random forest algorithm and the feature stack of the segmented part, and one or more phases in the segmented part are distinguished in the result of the visual phase segmentation.

6. The electronic device according to claim 5, characterized in that: The electronic device further comprises: A conversion module, used for converting the result of the visual phase segmentation into a binary image; A first calculation module, configured to calculate the volume fraction of the one or more phases based on the area fraction of the one or more phases in the binary image; The second calculation module is used to calculate the porosity in the composite cement system and the hydration degree of the multiple cementitious materials based on the volume fraction.

7. The electronic device according to claim 6, characterized in that: The first random forest algorithm is also used to perform visual phase segmentation on phases of specified sizes and / or phases of specified classification in the composite cement system.

8. The electronic device according to claim 7, characterized in that: The different feature extraction methods include Gaussian filtering, Sobel filtering and pixel averaging.

9. An electronic device comprising a memory, at least one 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 method according to any one of claims 1 to 4 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 4 is implemented.

Images