Image three-dimensional reconstruction method, device, equipment and medium

Abstract

The present invention relates to the field of image processing technology, and in particular to a method, device, equipment and medium for three-dimensional reconstruction of an image, the method comprising: obtaining a first electron microscope scanning image of a gas diffusion layer in a proton exchange membrane fuel cell; inputting the first electron microscope scanning image into a pre-constructed image recognition model component to obtain the components of each first regional block in the first electron microscope scanning image; restoring a single-layer carbon fiber structure based on the components of each first regional block in the first electron microscope scanning image and the thickness of the first regional block; reconstructing a three-dimensional gas diffusion layer based on the single-layer carbon fiber structure, the three-dimensional gas diffusion layer comprising a single-layer carbon fiber structure and a multi-layer regional block structure. The implementation of the present invention can ensure an accurate reconstruction of a three-dimensional gas diffusion layer, and then the corresponding performance conditions calculated through the three-dimensional gas diffusion layer are more accurate, thereby promoting the rapid development of hydrogen energy batteries.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a method, device, equipment and medium for three-dimensional image reconstruction. Background Art

[0002] In recent years, the development of the hydrogen energy industry has received much attention from countries around the world, and hydrogen energy extraction technology and fuel cell manufacturing technology have also achieved breakthrough developments.

[0003] Hydrogen energy cells include a gas diffusion layer or diffusion layer, whose main functions include conducting current, transporting gaseous reactants, discharging liquid water, supporting the catalyst layer, and realizing the redistribution of gaseous reactants between the flow field and the catalyst layer. The diffusion layer is the thickest component in the porous electrode, so the gas diffusion layer can be regarded as the main part of the porous electrode and is one of the key components affecting the performance of the electrode. The gas diffusion layer is composed of conductive porous materials, commonly used materials are carbon paper or carbon cloth, and there are also non-woven fabrics and carbon black paper. A good gas diffusion layer should meet the following requirements: good air permeability, high electronic conductivity, compact structure and flat surface to reduce the contact resistance generated by assembly, uniform porous structure, certain mechanical strength, appropriate rigidity and flexibility are conducive to the manufacture of electrodes, good chemical stability and thermal stability and low manufacturing cost. Transporting gas and discharging liquid water are important functions of the diffusion layer.

[0004] The existing method for calculating the parameters of the gas diffusion layer uses the electron microscope scanning method. The scanning electron microscope image is used to observe and analyze the three-dimensional morphology of the gas diffusion layer. While observing the morphology, the composition analysis of the micro-area is performed manually. It cannot characterize the real structure of the gas diffusion layer, and the calculation results are not accurate. Summary of the invention

[0005] In order to solve the technical problems existing in the prior art, that is, the gas diffusion layer cannot accurately restore the three-dimensional structure and the calculation problem based on the existing gas diffusion layer performance is inaccurate, the present invention provides a three-dimensional image reconstruction method, the method comprising:

[0006] Acquire a first electron microscope scanning image of a gas diffusion layer in a proton exchange membrane fuel cell;

[0007] Inputting the first electron microscope scanning image into a pre-constructed image recognition model component to obtain components of each first regional block in the first electron microscope scanning image, wherein the image recognition model component is used to divide the first electron microscope scanning image into a plurality of interconnected first regional blocks and predict the components of each first regional block;

[0008] Restoring a single-layer carbon fiber structure based on the components of each first regional block in the first electron microscope scanning image and the thickness of the first regional block;

[0009] A three-dimensional gas diffusion layer is reconstructed based on the single-layer carbon fiber structure, wherein the three-dimensional gas diffusion layer includes a single-layer carbon fiber structure and a multi-layer regional block structure, each layer of the regional block structure includes a second regional block corresponding to the first regional block and components of each of the second regional blocks, and the difference between the thickness of the second regional block and the thickness of the first regional block is less than a preset thickness threshold.

[0010] Furthermore, restoring the single-layer carbon fiber structure based on the components of each first region block in the first electron microscope scanning image and the thickness of the first region block includes:

[0011] Determining the area ratio corresponding to each component according to the number of first regional blocks corresponding to each component in the first electron microscope scanning image;

[0012] Determining the thickness of the single-layer carbon fiber structure based on the area ratios corresponding to the respective components;

[0013] The single-layer carbon fiber structure is restored based on the thickness of the single-layer carbon fiber structure.

[0014] Furthermore, before reconstructing the three-dimensional gas diffusion layer based on the single-layer carbon fiber structure, the method further includes:

[0015] Adding a component label to the corresponding first regional block based on the components of each first regional block;

[0016] The method of reconstructing a three-dimensional gas diffusion layer based on the single-layer carbon fiber structure comprises:

[0017] Determining the number of layers of the three-dimensional gas diffusion layer based on a preset thickness and the thickness of each of the first region blocks;

[0018] Taking the single-layer carbon fiber structure as the first-layer carbon fiber structure, laying multiple layers of regional block structures on the first-layer carbon fiber structure in sequence until the sum of the number of layers of the regional block structure and the number of layers of the single-layer carbon fiber structure reaches the number of layers of the three-dimensional gas diffusion layer, wherein the regional block structure includes a plurality of second regional blocks to which component labels are not added;

[0019] According to the preset ratio requirements of each component, component labels are added to the second area blocks in each area block structure respectively, so that each component in each layer of the area block structure meets the preset ratio requirements, thereby obtaining the three-dimensional gas diffusion layer.

[0020] Furthermore, the method further comprises:

[0021] Acquire the number of the first region blocks corresponding to each component in the three-dimensional gas diffusion layer, the number of the second region blocks corresponding to each component, and the density corresponding to each component;

[0022] Calculating a first mass of a real three-dimensional gas diffusion layer corresponding to the three-dimensional gas diffusion layer based on the number of the first region blocks corresponding to each component, the number of the second region blocks corresponding to each component, and the density corresponding to each component;

[0023] Acquire a second mass of the real three-dimensional gas diffusion layer corresponding to the first electron microscope scanning image;

[0024] When the difference between the first mass and the second mass is less than a preset mass threshold, it is determined that the three-dimensional gas diffusion layer is a three-dimensional point cloud corresponding to the real three-dimensional gas diffusion layer.

[0025] Furthermore, before inputting the first electron microscope scanning image into a pre-built image recognition model component to obtain the components of each first regional block in the first electron microscope scanning image, the method includes:

[0026] Acquire a first diameter of the carbon fiber in the first electron microscope scanning image and a second diameter of the carbon fiber of the real three-dimensional gas diffusion layer corresponding to the first electron microscope scanning image;

[0027] Determine and scale the first electron microscope scanning image based on the ratio of the first diameter to the second diameter to obtain a second electron microscope scanning image;

[0028] Accordingly, the first electron microscope scanning image is input into a pre-built image recognition model component to obtain components of each first region block in the first electron microscope scanning image, including:

[0029] The second electron microscope scanning image is input into a pre-constructed image recognition model component to obtain the components of each first area block in the second electron microscope scanning image.

[0030] Furthermore, according to the preset ratio requirements of each component, component labels are added to the second area blocks in each area block structure respectively, so that each component in each layer of the area block structure meets the preset ratio requirements, including:

[0031] Selecting target layers in the multi-layer region block structure in sequence according to a preset order;

[0032] Randomly select two second area blocks in the selected target layer, and mark the second area block between the two selected second area blocks as a carbon fiber label;

[0033] Repeat the step of randomly selecting two second area blocks in the selected target layer and marking the second area block between the two selected second area blocks as a carbon fiber label until the area ratio of the carbon fiber in the target layer reaches a preset area ratio;

[0034] Obtaining a node area block where a node of the carbon fiber in the target layer is located, marking an unmarked second area block around the node area block as an adhesive label, wherein the node of the carbon fiber is an intersection point of continuous carbon fiber labels in different directions;

[0035] The unmarked second area block around the adhesive label is marked as a hydrophobic agent label.

[0036] Furthermore, after marking the unmarked second area block around the adhesive label as a hydrophobic agent label, the method further includes:

[0037] reselecting a target layer in the multi-layer region block structure;

[0038] Repeat the step of randomly selecting two second area blocks in the selected target layer, and marking the second area block between the two selected second area blocks as a carbon fiber label.

[0039] In another aspect, the present invention provides a three-dimensional image reconstruction device, the device comprising:

[0040] An image acquisition module, used to acquire a first electron microscope scanning image of a gas diffusion layer in a proton exchange membrane fuel cell;

[0041] a component determination module, used for inputting the first electron microscope scanning image into a pre-constructed image recognition model component to obtain the components of each first regional block in the first electron microscope scanning image, wherein the image recognition model component is used for dividing the first electron microscope scanning image into a plurality of interconnected first regional blocks and predicting the components of each first regional block;

[0042] A restoration module, used to restore a single-layer carbon fiber structure based on the components of each first region block in the first electron microscope scanning image and the thickness of the first region block;

[0043] A three-dimensional reconstruction module is used to reconstruct a three-dimensional gas diffusion layer based on the single-layer carbon fiber structure, wherein the three-dimensional gas diffusion layer includes a single-layer carbon fiber structure and a multi-layer regional block structure, each layer of the regional block structure includes a second regional block corresponding to the first regional block and a component of each second regional block, and the difference between the thickness of the second regional block and the thickness of the first regional block is less than a preset thickness threshold.

[0044] In another aspect, the present invention provides an electronic device, comprising:

[0045] processor;

[0046] a memory for storing instructions executable by the processor;

[0047] Wherein, the processor is configured to execute the instructions to implement the three-dimensional image reconstruction method as described above.

[0048] On the other hand, the present invention provides a computer-readable storage medium. When the instructions in the computer-readable storage medium are executed by a processor of an image 3D reconstruction device / electronic device, the image 3D reconstruction device / electronic device can perform the image 3D reconstruction as described above.

[0049] The implementation of this application has the following beneficial effects:

[0050] The three-dimensional image reconstruction method, device, equipment and medium provided in the present application can obtain the first electron microscope scanning image of the corresponding gas diffusion layer through electron microscope scanning, and identify the first regional block where each component in the first electron microscope scanning image is located through a pre-built image recognition model component, so that each component can be identified more accurately. At the same time, thickness can be added to each first regional block, and the thickness can restore the two-dimensional first regional block to a three-dimensional regional block, thereby realizing the restoration of the single-layer carbon fiber structure in the first electron microscope scanning image, and then a three-dimensional gas diffusion layer with a user-set thickness can be reconstructed on the basis of the single-layer carbon fiber structure. The three-dimensional gas diffusion layer is more in line with the distribution and proportion of components such as carbon fibers in the actual gas diffusion layer, ensuring that the corresponding performance conditions calculated by the reconstructed three-dimensional gas diffusion layer are more accurate, thereby promoting the rapid development of hydrogen energy batteries. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required for use in the embodiments or prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention, and for those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work.

[0052] Figure 1 It is a flowchart of a method for three-dimensional image reconstruction provided by an embodiment of the present application;

[0053] Figure 2 is a schematic diagram of a first electron microscope scanning image provided in an embodiment of the present application;

[0054] Figure 3 It is a flowchart of another method for three-dimensional image reconstruction provided in an embodiment of the present application;

[0055] Figure 4is a structural schematic diagram of a three-dimensional image reconstruction device provided by an embodiment of the present invention;

[0056] Figure 5 It is a structural schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0057] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0058] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, device, product or equipment that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or equipment.

[0059] In order to better understand this application, the following technical terms are explained:

[0060] In order to implement the technical solution of the present application and enable more engineering and technical workers to easily understand and apply the present application, the working principle of the present application will be further explained in combination with specific embodiments.

[0061] The following describes an embodiment of a method for three-dimensional image reconstruction according to the present application. Figure 1 This is a flowchart of a method for three-dimensional image reconstruction provided by an embodiment of the present application. This specification provides method operation steps as described in the embodiment or flowchart, but based on routine or non-creative work, more or fewer operation steps may be included. The order of steps listed in the embodiment is only one way of executing the steps among many, and does not represent the only execution order. Specifically, Figure 1 As shown, the execution subject of the scheme may be a terminal, and the terminal may communicate with an electron microscope scanning instrument. The method may include:

[0062] S102, obtaining a first electron microscope scanning image of a gas diffusion layer in a proton exchange membrane fuel cell.

[0063] Specifically, the components of the gas diffusion layer mainly include carbon fiber, adhesive (phenolic resin), hydrophobic agent (polytetrafluoroethylene) and some other additives.

[0064] The first electron microscope scanning image can be an electron microscope image of a gas diffusion layer in a proton exchange membrane fuel cell collected by the electron microscope scanning instrument, and the first electron microscope scanning image can include images of the above components. The first electron microscope scanning image can also include gaps formed by the carbon fiber structure.

[0065] like Figure 2 As shown, Figure 2 It is a schematic diagram of a first electron microscope scanning image provided in an embodiment of the present application. It can be seen that the first electron microscope scanning image is for the upper surface or the lower surface of the gas diffusion layer. It can be understood that the gas diffusion layer is formed by bonding multiple layers of carbon fibers with an adhesive. Accordingly, images corresponding to multiple layers of carbon fibers can be collected in the first electron microscope scanning image.

[0066] S104, inputting the first electron microscope scanning image into a pre-constructed image recognition model component to obtain components of each first regional block in the first electron microscope scanning image.

[0067] In an optional embodiment, before reconstructing the three-dimensional gas diffusion layer based on the single-layer carbon fiber structure, the method further includes: adding a component label to the corresponding first regional block based on the component of each first regional block.

[0068] Specifically, the image recognition model component is used to divide the first electron microscope scanning image into multiple interconnected first area blocks, and mark the component labels corresponding to the components of each first area block. The first area block contains only one component, and the components include: carbon fiber, hydrophobic agent, adhesive and gap. The component labels include carbon fiber labels, hydrophobic agent labels, adhesive labels and gap labels.

[0069] Specifically, the image recognition model component is established according to the following method:

[0070] Acquire multiple sets of sample electron microscope scanning images, and mark the boundary information of each component in the sample electron microscope scanning images;

[0071] Specifically, the first electron microscope scanning image can be classified and feature extracted to determine the boundary information of each component in the first electron microscope scanning image. The purpose of classification is to train image recognition model components for samples of different sample electron microscope scanning images. In the early stage of feature extraction, the sample electron microscope scanning image samples can be processed manually to divide the boundary information of carbon fiber, adhesive and hydrophobic agent.

[0072] The sample electron microscope scanning image is divided into a plurality of first area blocks.

[0073] The component of each first region block is determined based on the boundary information of each component.

[0074] Specifically, the second electron microscope scanning image can be divided according to the diameter of the carbon fiber to obtain a plurality of first area blocks with the same area. Each first area block has only one component.

[0075] In an optional embodiment, when there are more than two components corresponding to the first regional block, the components of the first regional block can be determined according to the area of ​​each component, that is, the component with the largest area in the first regional block is used as the component of the first regional block.

[0076] The image recognition model component is established, wherein the image recognition model component includes a plurality of model parameters.

[0077] The multiple groups of second electron microscope scanning images are used as input data of the image recognition model component, the components of each corresponding first area block are used as output data of the image recognition model component, and the model parameters of the image recognition model component are adjusted until the image recognition model component meets the preset requirements.

[0078] Specifically, UNet semantic segmentation technology can be used in the image recognition model component to identify the boundary information of each component.

[0079] In an optional embodiment, the image recognition model component may include an input layer: for inputting a second electron microscope scanning image; a convolution layer: using a convolution kernel to extract each component and perform feature mapping; an excitation layer (activation function): since convolution is also a linear operation, it is necessary to add nonlinear mapping, wherein the Relu function is used as the activation function in the embodiment of this specification; a pooling layer: for downsampling, sparse processing of the feature map, and reducing the amount of data calculation; a fully connected layer: for refitting to reduce the loss of feature information.

[0080] Specifically, an image recognition model component can be established, and the image recognition model component includes multiple model parameters, and the model parameters can represent constraints, which can be specifically set according to expert experience, etc. The above-mentioned multiple groups of second electron microscope scanning images can be used to perform model training on the image recognition model component, and the second electron microscope scanning image can include boundary information of each component or components corresponding to each first regional block. Each component in multiple groups of second electron microscope scanning images is used as the input of the image recognition model component, and the components corresponding to each first regional block corresponding to the second electron microscope scanning image are used as output, and the model parameters in the image recognition model component are continuously adjusted until the image recognition model component reaches a preset requirement such as: meeting the preset accuracy, and the number of model parameter adjustments meets the preset number of requirements, then the model training is completed. Specifically, machine learning algorithms such as: GBDT (Gradient Boosting Decision Tree, gradient boosting decision tree) algorithm can be referenced for model training.

[0081] Model training is performed based on multiple sets of second electron microscope scanning images to construct an image recognition model component, which can identify components that are not significant in the scanning electron microscope image, thereby improving the accuracy of component determination. After the first electron microscope scanning image is input into the trained image recognition model component, the components of each first area block in the first electron microscope scanning image can be output.

[0082] S106: Restore a single-layer carbon fiber structure based on the components of each first region block in the first electron microscope scanning image and the thickness of the first region block.

[0083] Specifically, the thickness of the first area block can be specified by the user or determined according to the diameter of the carbon fiber. The thickness of the first area block should be greater than the diameter of the carbon fiber. After determining the thickness of the first area block, each first area block can be restored to a three-dimensional area block, that is, the first area block is two-dimensional, and the restored single-layer carbon fiber structure includes a three-dimensional area block corresponding to each first area block.

[0084] S108. Reconstructing a three-dimensional gas diffusion layer based on the single-layer carbon fiber structure, wherein the three-dimensional gas diffusion layer includes a single-layer carbon fiber structure and a multi-layer regional block structure.

[0085] Specifically, each layer of the regional block structure includes a second regional block corresponding to the first regional block and components of each of the second regional blocks, and a difference between a thickness of the second regional block and a thickness of the first regional block is less than a preset thickness threshold.

[0086] Specifically, the three-dimensional gas diffusion layer can be understood as a real gas diffusion layer, which is composed of multiple layers of carbon fiber structures. The thickness of the three-dimensional gas diffusion layer can correspond to the real gas diffusion layer, that is, if the thickness of the real gas diffusion layer is 10um, then the three-dimensional gas diffusion layer is also 10um. The thickness of different regional block structures can be different from the thickness of the single-layer carbon fiber structure, and the difference between the two can be less than the preset thickness threshold, wherein the preset thickness threshold is not specifically limited in the embodiments of this specification and can be set according to actual needs.

[0087] The three-dimensional image reconstruction method provided in the embodiments of the present specification can obtain the first electron microscope scanning image of the corresponding gas diffusion layer through electron microscope scanning, and identify the first area block where each component in the first electron microscope scanning image is located through a pre-built image recognition model component, so that the identification of each component is more accurate, and at the same time, thickness can be added to each first area block, and the thickness can restore the two-dimensional first area block to a three-dimensional area block, thereby realizing the restoration of the single-layer carbon fiber structure in the first electron microscope scanning image, and at the same time, it can also ensure that the restored single-layer carbon fiber structure is more accurate and reduces errors. Then, based on the single-layer carbon fiber structure, a three-dimensional gas diffusion layer with a user-set thickness can be reconstructed, and the three-dimensional gas diffusion layer is more in line with the distribution and proportion of components such as carbon fibers in the actual gas diffusion layer, ensuring that the corresponding performance conditions calculated by the reconstructed three-dimensional gas diffusion layer are more accurate, thereby promoting the rapid development of hydrogen energy batteries.

[0088] Based on the above embodiment, in one embodiment of the present specification, the method of restoring a single-layer carbon fiber structure based on the components of each first region block in the first electron microscope scanning image and the thickness of the first region block includes:

[0089] Determining the area ratio corresponding to each component according to the number of first regional blocks corresponding to each component in the first electron microscope scanning image;

[0090] Specifically, the embodiments of the present specification can determine the area of ​​each component by counting the number of components of different first area blocks, and according to the number of each component, for example, the area of ​​the first area block is 1 unit, for example, 1 unit is 1 square centimeter, wherein the numbers of carbon fibers, hydrophobic agents, adhesives and gaps are 45, 10, 15, and 30 respectively, wherein the areas corresponding to the single numbers are the same, that is, the areas corresponding to a number of carbon fibers, a number of hydrophobic agents, a number of adhesives, and a number of gaps are the same, then the areas corresponding to the carbon fibers, hydrophobic agents, adhesives, and gaps are 0.45 square centimeters, 0.10 square centimeters, 0.15 square centimeters, and 0.30 square centimeters respectively, that is, the area ratios corresponding to the carbon fibers, hydrophobic agents, adhesives, and gaps are 45%, 10%, 15%, and 30% respectively.

[0091] The thickness of the single-layer carbon fiber structure is determined based on the area ratios corresponding to the various components.

[0092] The single-layer carbon fiber structure is restored based on the thickness of the single-layer carbon fiber structure.

[0093] Specifically, the area ratios of the components in the single-layer carbon fiber structure are different, and the corresponding thicknesses are also different. For example, the area ratios of carbon fiber, hydrophobic agent, adhesive and gap are 45%, 10%, 15% and 30%, respectively. By looking up the table, the corresponding thickness can be determined to be 2um; for example, the area ratios of carbon fiber, hydrophobic agent, adhesive and gap are 30%: 13%: 11%: 46%, respectively. By looking up the table, the corresponding thickness can be determined to be 3um. That is, the thickness of the single-layer carbon fiber structure corresponding to the different area ratios of the embodiment of this specification can establish a corresponding relationship in advance, that is, the different area ratios of each component can correspond to different thicknesses of the single-layer carbon fiber structure. In practical applications, the thickness of the single-layer carbon fiber structure can be determined according to the area ratio by looking up the table. In an optional embodiment, the thickness of the single-layer carbon fiber structure can be determined according to the area ratio of the carbon fiber. For example, if the area ratio of the carbon fiber is 20%, the thickness of the single-layer carbon fiber structure is 2um, and if the area ratio of the carbon fiber is 30%, the thickness of the single-layer carbon fiber structure is 3um. The above table may store the area ratio of different components and the corresponding thickness relationship, that is, the table may be pre-established, such as the thickness is 3, the corresponding area ratio may be 30%: 13%: 11%: 46%, 30%: 20%: 10%: 40%, 20%: 20%: 15%: 45%, etc. In practical applications, the corresponding thickness may be determined by directly comparing the table.

[0094] The three-dimensional image reconstruction method provided in the embodiment of this specification can determine the thickness of a single-layer carbon fiber structure by the area ratio of each component, which is more in line with the generation process of the gas diffusion layer. At the same time, the three-dimensional gas diffusion layer restored using the thickness of the single-layer carbon fiber structure is more accurate, thereby improving the accuracy of the three-dimensional gas diffusion layer reconstruction.

[0095] Based on the above embodiments, in one embodiment of this specification, Figure 3 is a flow chart of another method for three-dimensional image reconstruction provided by an embodiment of the present application, such as Figure 3 As shown, the three-dimensional gas diffusion layer having a preset thickness is reconstructed based on the single-layer carbon fiber structure, including:

[0096] S302: Determine the number of layers of the three-dimensional gas diffusion layer based on a preset thickness and the thickness of each of the first region blocks.

[0097] Specifically, the preset thickness may be specified by the user. In practical applications, after the user specifies the thickness, the preset thickness may be divided by the thickness of the first region block to determine the number of layers of the three-dimensional gas diffusion layer. For example, if the preset thickness is 20um and the thickness of the first region block is 2um, the number of layers of the three-dimensional gas diffusion layer may be determined to be 10. When the preset thickness cannot divide the thickness of the first region block, the number of layers of the three-dimensional gas diffusion layer may be determined by adding one layer. For example, if the preset thickness is 20um and the thickness of the first region block is 3um, the number of layers of the three-dimensional gas diffusion layer may be determined to be 7.

[0098] S304. Taking the single-layer carbon fiber structure as the first-layer carbon fiber structure, a plurality of regional block structures are sequentially laid on the first-layer carbon fiber structure until the sum of the number of layers of the regional block structure and the number of layers of the single-layer carbon fiber structure reaches the number of layers of the three-dimensional gas diffusion layer, wherein the regional block structure includes a plurality of second regional blocks to which component labels are not added.

[0099] Specifically, the second area block may have thickness, and each second area block may correspond to a first area block one by one, that is, the second area block may be stacked on the corresponding three-dimensional first area block.

[0100] S306 , according to the preset ratio requirement of each component, respectively add component labels to the second region blocks in each region block structure, so that each component in each layer of the region block structure meets the preset ratio requirement, thereby obtaining the three-dimensional gas diffusion layer.

[0101] In an optional embodiment, according to the preset ratio requirement of each component, adding component labels to the second regional blocks in each of the regional block structures respectively so that each component in each layer of the regional block structure meets the preset ratio requirement includes:

[0102] In the multi-layer region block structure, target layers are selected in sequence according to a preset order.

[0103] Specifically, a multi-layer region block structure can be used as a set of target layers to be selected, and the set includes region block structures of each layer. When selecting a target layer, a region block structure corresponding to any subset in the set can be selected as the target layer, and the subset can be deleted from the set after selection.

[0104] Specifically, the preset order may be the order of the subsets in the above set.

[0105] Two second area blocks are randomly selected in the selected target layer, and the second area block between the two selected second area blocks is marked as a carbon fiber label.

[0106] Specifically, the target layer includes multiple second area blocks, each of which can be stored in the form of a vector matrix. In practical applications, any two second area blocks can be selected from the vector matrix, and the second area block between the two selected second area blocks can be marked as a carbon fiber label.

[0107] Repeat the step of randomly selecting two second area blocks in the selected target layer and marking the second area block between the two selected second area blocks as a carbon fiber label until the area ratio of the carbon fiber in the target layer reaches a preset area ratio.

[0108] Specifically, the area ratio of the carbon fiber of the target layer is determined based on the number of second area blocks with carbon fiber labels. The preset area ratio can be set by the user or can be the area ratio of the carbon fiber in a single-layer carbon fiber structure.

[0109] A node area block where the node of the carbon fiber in the target layer is located is obtained, and an unlabeled second area block around the node area block is marked as an adhesive label, wherein the node of the carbon fiber is an intersection point of continuous carbon fiber labels in different directions.

[0110] Specifically, the node area block can be understood as having at least two second area blocks with carbon fiber labels around the corresponding second area block. It can be understood that the second area blocks around the second area block are directly connected. In practical applications, the unlabeled second area blocks around the node area block can be labeled as adhesive labels. The above steps can be understood as arranging adhesives at the carbon fiber nodes, randomly arranging them according to the characteristics of the first layer, and generating a three-dimensional structure according to the predicted thickness until the adhesive ratio reaches the first layer distribution ratio.

[0111] The unmarked second area block around the adhesive label is marked as a hydrophobic agent label.

[0112] In an optional embodiment, after marking the unmarked second area block around the adhesive label as a hydrophobic agent label, the method further includes:

[0113] reselecting a target layer in the multi-layer region block structure;

[0114] Repeat the step of randomly selecting two second area blocks in the selected target layer, and marking the second area block between the two selected second area blocks as a carbon fiber label.

[0115] In an optional embodiment, the method further comprises: adding a gap label to the second region block to which no label is added.

[0116] The embodiments of this specification are based on the correlation between the surface characteristics and three-dimensional characteristics of the gas diffusion layer during the hot pressing process. By accurately identifying and summarizing the distribution and growth laws of multiple components on the surface of the fuel cell gas diffusion layer, it is possible to establish a restoration of the three-dimensional structure of the entire gas diffusion layer to ensure the accuracy of the restoration.

[0117] Based on the above embodiment, in one embodiment of this specification, the method further includes:

[0118] The number of the first region blocks corresponding to each component in the three-dimensional gas diffusion layer, the number of the second region blocks corresponding to each component, and the density corresponding to each component are obtained.

[0119] Specifically, the density corresponding to each component can be preset, such as: the density of air is 0kg / m3, the carbon fiber and the adhesive are both carbon with a density of 1kg / m3, and the density of the hydrophobic agent is 3kg / m3.

[0120] The first mass of a real three-dimensional gas diffusion layer corresponding to the three-dimensional gas diffusion layer is calculated based on the number of the first region blocks corresponding to each component, the number of the second region blocks corresponding to each component, and the density corresponding to each component.

[0121] Acquire a second mass of the real three-dimensional gas diffusion layer corresponding to the first electron microscope scanning image;

[0122] When the difference between the first mass and the second mass is less than a preset mass threshold, it is determined that the three-dimensional gas diffusion layer is a three-dimensional point cloud corresponding to the real three-dimensional gas diffusion layer.

[0123] Specifically, the three-dimensional gas diffusion layer of preset thickness is V, and there are n regional blocks in the single-layer carbon fiber structure and the multi-layer regional block structure, and the regional blocks include the first regional block and the second regional block. Based on the number of regional blocks corresponding to each component, the density of the three-dimensional gas diffusion layer of preset thickness can be determined to be (n0*ρ0+(n1+n2)*ρ1+n3*ρ3) / n, so the overall model quality is ((n*ρ0+(n1+n2)*ρ1+n3*ρ3) / n)*V / n. Among them, n0 is the number of gap labels, ρ0 is the air density or gap density, n1 is the number of carbon fiber labels, n2 is the number of adhesive labels, ρ1 is the density of carbon fiber and adhesive, n3 is the number of hydrophobic agent labels, ρ3 is the hydrophobic agent density, and n is the total number of regional blocks. In practical applications, the preset quality threshold can be understood as being pre-set. For example, if the difference between the mass of the three-dimensional gas diffusion layer and the mass of the real three-dimensional gas diffusion layer is not less than 1% of the mass of the real three-dimensional gas diffusion layer, it is a qualified output, and the corresponding point cloud of each area block can be output. It can be understood that the three-dimensional point cloud can be understood as the above-mentioned vector matrix, that is, each area block corresponds to a data in the vector matrix. The three-dimensional point cloud is used to distinguish and determine the correspondence between different area blocks. For example, the three-dimensional point cloud corresponding to the three-dimensional gas diffusion layer is a 10 (length) * 10 (width) * 9 (height) cuboid, then the area block can be represented by (a, b, c), such as a corresponds to the layer, b corresponds to the length, and c corresponds to the width.

[0124] The embodiments of the present specification can compare the quality of the restored three-dimensional gas diffusion layer and the real gas diffusion layer. After the comparison, the three-dimensional point cloud corresponding to each regional block in the restored three-dimensional gas diffusion layer can be output, thereby improving the accuracy of the three-dimensional point cloud output and ensuring the accuracy of the gas diffusion layer performance calculated based on the three-dimensional point cloud.

[0125] Based on the above embodiment, in one embodiment of the present specification, before inputting the first electron microscope scanning image into a pre-built image recognition model component to obtain the components of each first regional block in the first electron microscope scanning image, the method includes:

[0126] Acquire a first diameter of the carbon fiber in the first electron microscope scanning image and a second diameter of the carbon fiber of the real three-dimensional gas diffusion layer corresponding to the first electron microscope scanning image;

[0127] Determine and scale the first electron microscope scanning image based on the ratio of the first diameter to the second diameter to obtain a second electron microscope scanning image;

[0128] Specifically, in order to better calculate the components corresponding to each first area block, the first electron microscope scanning image can be proportionally reduced or enlarged according to the ratio of the first diameter to the second diameter. It can be understood that in the input first electron microscope scanning image, since the distance between the electron microscope scanning and the gas diffusion layer in the proton exchange membrane fuel cell is different, the proportions of different first electron microscope scanning images scanned will be different. In practical applications, when the ratio of the first diameter to the second diameter is greater than one, the first electron microscope scanning image can be reduced, and when the ratio of the first diameter to the second diameter is less than one, the first electron microscope scanning image can be proportionally enlarged.

[0129] Accordingly, the first electron microscope scanning image is input into a pre-built image recognition model component to obtain components of each first region block in the first electron microscope scanning image, including:

[0130] The second electron microscope scanning image is input into a pre-constructed image recognition model component to obtain the components of each first area block in the second electron microscope scanning image.

[0131] The three-dimensional image reconstruction method provided in the embodiment of the present specification determines the scaling ratio of the first electron microscope scanning image by using the first diameter of the carbon fiber in the first electron microscope scanning image and the second diameter of the carbon fiber of the real three-dimensional gas diffusion layer corresponding to the first electron microscope scanning image, thereby obtaining an image consistent with the size of the real gas diffusion layer, and identifies the components of each first area block in the corresponding second electron microscope scanning image through an image recognition model component, thereby ensuring more accurate identification and determination of each component, thereby avoiding errors in gas diffusion layer reconstruction caused by inconsistent scaling ratios of the electron microscope scanning instrument and excessive deviations in calculated performance indicators.

[0132] The present invention is based on the association between the surface characteristics and three-dimensional characteristics of the gas diffusion layer during hot pressing. Through the accurate identification and induction of the distribution and growth laws of multiple components on the surface of the fuel cell gas diffusion layer, the restoration of the three-dimensional structure of the entire gas diffusion layer can be established. The implementation of this method can quickly identify multiple components of the finished gas diffusion layer, and at the same time establish a three-dimensional structure of the gas diffusion layer that is consistent with the actual situation, and timely feedback the growth law of the components to the gas diffusion layer manufacturer, adjust the product production material ratio based on this information, and guide the gas diffusion layer and other porous media related material manufacturing companies to better plan process flow and manufacturing plans. At the same time, it provides accurate gas diffusion layer microstructure parameters, such as carbon fiber, adhesive, hydrophobic agent, pore distribution and generation law, for scientific researchers, and at the same time improves the accuracy of the reconstructed model and the multi-component status, making the simulation research closer to the real situation.

[0133] The above method can be applied to the fuel cell gas diffusion layer industry and other carbon paper porous media industries. It can solve the research and manufacturing problems of porous media in the fields of petrochemicals, air purification, textiles, papermaking, aerospace, atomic energy, etc., and has great scientific significance and industrial value.

[0134] In another aspect, the present invention provides a three-dimensional image reconstruction device. Figure 4 is a schematic diagram of a structure of a three-dimensional image reconstruction device provided by an embodiment of the present invention, referring to Figure 4 , the device may include:

[0135] An image acquisition module 401 is used to acquire a first electron microscope scanning image of a gas diffusion layer in a proton exchange membrane fuel cell;

[0136] A component determination module 402 is used to input the first electron microscope scanning image into a pre-built image recognition model component to obtain the components of each first regional block in the first electron microscope scanning image. The image recognition model component is used to divide the first electron microscope scanning image into a plurality of interconnected first regional blocks and predict the components of each first regional block.

[0137] A restoration module 403 is used to restore a single-layer carbon fiber structure based on the components of each first region block in the first electron microscope scanning image and the thickness of the first region block;

[0138] The three-dimensional reconstruction module 404 is used to reconstruct a three-dimensional gas diffusion layer based on the single-layer carbon fiber structure, wherein the three-dimensional gas diffusion layer includes a single-layer carbon fiber structure and a multi-layer regional block structure, each layer of the regional block structure includes a second regional block corresponding to the first regional block and a component of each second regional block, and the difference between the thickness of the second regional block and the thickness of the first regional block is less than a preset thickness threshold.

[0139] It should be noted that the device provided in the above embodiment, when implementing its functions, is only illustrated by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the device and method embodiments provided in the above embodiment belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0140] on the other hand, Figure 5 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention, such as Figure 5As shown, the present invention provides an electronic device for the image three-dimensional reconstruction method, the device includes a processor and a memory, the memory stores at least one instruction or at least one program, the at least one instruction or at least one program is loaded and executed by the processor to implement the image three-dimensional reconstruction method as described above.

[0141] If the above functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention.

[0142] An embodiment of the present invention also provides a computer-readable storage medium, which stores at least one instruction, at least one program, a code set or an instruction set. The at least one instruction, the at least one program, the code set or the instruction set can be executed by a processor of an electronic device to complete the above-mentioned image three-dimensional reconstruction method.

[0143] Optionally, in an embodiment of the present invention, the above-mentioned storage medium may include but is not limited to: a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and other media that can store program codes.

[0144] It should be noted that the sequence of the embodiments of the present invention described above is for description only and does not represent the advantages and disadvantages of the embodiments. The above describes specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0145] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0146] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, electronic device and storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments.

[0147] A person skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware or by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.

[0148] It should be noted that each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. The test method provided in the embodiment of the present invention has the same implementation principle and technical effects as the aforementioned system embodiment. For the sake of brief description, for the parts not mentioned in the method embodiment, reference can be made to the corresponding contents in the aforementioned system embodiment.

[0149] In several embodiments provided in the present application, it should be understood that the disclosed system and method can also be implemented in other ways. The device embodiments described above are merely schematic. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the devices, methods and computer program products according to multiple embodiments of the present invention. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0150] Finally, it should be noted that the above-described embodiments are only specific implementations of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The protection scope of the present invention is not limited thereto. Although the present invention is described in detail with reference to the above-mentioned embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-mentioned embodiments within the technical scope disclosed by the present invention, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the above-mentioned claims.

Claims

1. A method for three-dimensional image reconstruction, characterized in that: The method comprises: Acquire a first electron microscope scanning image of a gas diffusion layer in a proton exchange membrane fuel cell; Inputting the first electron microscope scanning image into a pre-constructed image recognition model component to obtain components of each first regional block in the first electron microscope scanning image, wherein the image recognition model component is used to divide the first electron microscope scanning image into a plurality of interconnected first regional blocks and predict the components of each first regional block; Determining the area ratio corresponding to each component according to the number of first regional blocks corresponding to each component in the first electron microscope scanning image; Determining the thickness of the single-layer carbon fiber structure based on the area ratios corresponding to the respective components; Restoring the single-layer carbon fiber structure based on the thickness of the single-layer carbon fiber structure; Adding a component label to the corresponding first regional block based on the components of each first regional block; Determining the number of layers of the three-dimensional gas diffusion layer based on the preset thickness and the thickness of each of the first region blocks; Taking the single-layer carbon fiber structure as the first-layer carbon fiber structure, laying multiple layers of regional block structures on the first-layer carbon fiber structure in sequence until the sum of the number of layers of the regional block structure and the number of layers of the single-layer carbon fiber structure reaches the number of layers of the three-dimensional gas diffusion layer, wherein the regional block structure includes a plurality of second regional blocks to which component labels are not added; According to the preset proportion requirements of each component, component labels are added to the second regional blocks in each of the regional block structures, so that each component in each layer of the regional block structure meets the preset proportion requirements, and the three-dimensional gas diffusion layer is obtained, wherein the three-dimensional gas diffusion layer includes a single-layer carbon fiber structure and a multi-layer regional block structure, each layer of the regional block structure includes a second regional block corresponding to the first regional block and a component of each second regional block, the second regional block is stacked on the corresponding three-dimensional first regional block, and the difference between the thickness of the second regional block and the thickness of the first regional block is less than a preset thickness threshold.

2. The method for three-dimensional image reconstruction according to claim 1, characterized in that: The method further comprises: Acquire the number of the first region blocks corresponding to each component in the three-dimensional gas diffusion layer, the number of the second region blocks corresponding to each component, and the density corresponding to each component; Calculating a first mass of a real three-dimensional gas diffusion layer corresponding to the three-dimensional gas diffusion layer based on the number of the first region blocks corresponding to each component, the number of the second region blocks corresponding to each component, and the density corresponding to each component; Acquire a second mass of the real three-dimensional gas diffusion layer corresponding to the first electron microscope scanning image; When the difference between the first mass and the second mass is less than a preset mass threshold, it is determined that the three-dimensional gas diffusion layer is a three-dimensional point cloud corresponding to the real three-dimensional gas diffusion layer.

3. The method for three-dimensional image reconstruction according to claim 1, characterized in that: Before inputting the first electron microscope scanning image into a pre-built image recognition model component to obtain the components of each first regional block in the first electron microscope scanning image, the method includes: Acquire a first diameter of the carbon fiber in the first electron microscope scanning image and a second diameter of the carbon fiber of the real three-dimensional gas diffusion layer corresponding to the first electron microscope scanning image; Determine and scale the first electron microscope scanning image based on the ratio of the first diameter to the second diameter to obtain a second electron microscope scanning image; Accordingly, the first electron microscope scanning image is input into a pre-built image recognition model component to obtain components of each first region block in the first electron microscope scanning image, including: The second electron microscope scanning image is input into a pre-constructed image recognition model component to obtain the components of each first area block in the second electron microscope scanning image.

4. The method for three-dimensional image reconstruction according to claim 1, characterized in that: The adding component labels to the second regional blocks in each of the regional block structures according to the preset ratio requirements of each component, so that each component in each layer of the regional block structure meets the preset ratio requirements, includes: Selecting target layers in the multi-layer region block structure in sequence according to a preset order; Randomly select two second area blocks in the selected target layer, and mark the second area block between the two selected second area blocks as a carbon fiber label; Repeat the step of randomly selecting two second area blocks in the selected target layer and marking the second area block between the two selected second area blocks as a carbon fiber label until the area ratio of the carbon fiber in the target layer reaches a preset area ratio; Obtaining a node area block where a node of the carbon fiber in the target layer is located, marking an unmarked second area block around the node area block as an adhesive label, wherein the node of the carbon fiber is an intersection of continuous carbon fiber labels in different directions; The unmarked second area block around the adhesive label is marked as a hydrophobic agent label.

5. The method for three-dimensional image reconstruction according to claim 4, characterized in that: After marking the unmarked second area block around the adhesive label as a hydrophobic agent label, the method further includes: reselecting a target layer in the multi-layer region block structure; Repeat the step of randomly selecting two second area blocks in the selected target layer, and marking the second area block between the two selected second area blocks as a carbon fiber label.

6. A three-dimensional image reconstruction device, characterized in that: The device comprises: An image acquisition module, used to acquire a first electron microscope scanning image of a gas diffusion layer in a proton exchange membrane fuel cell; a component determination module, used for inputting the first electron microscope scanning image into a pre-constructed image recognition model component to obtain the components of each first regional block in the first electron microscope scanning image, wherein the image recognition model component is used for dividing the first electron microscope scanning image into a plurality of interconnected first regional blocks and predicting the components of each first regional block; An area ratio determination module, used to determine the area ratio corresponding to each component according to the number of first area blocks corresponding to each component in the first electron microscope scanning image; A thickness determination module, used to determine the thickness of the single-layer carbon fiber structure based on the area ratios corresponding to the respective components; A restoration module, used to restore the single-layer carbon fiber structure based on the thickness of the single-layer carbon fiber structure; A component label adding module, configured to add a component label to a corresponding first regional block based on the components of each first regional block; A layer number determination module, used to determine the number of layers of the three-dimensional gas diffusion layer based on a preset thickness and the thickness of each of the first regional blocks; A laying module, used for taking the single-layer carbon fiber structure as the first-layer carbon fiber structure, and sequentially laying a plurality of regional block structures on the first-layer carbon fiber structure, until the sum of the number of layers of the regional block structure and the number of layers of the single-layer carbon fiber structure reaches the number of layers of the three-dimensional gas diffusion layer, wherein the regional block structure includes a plurality of second regional blocks to which component labels are not added; A three-dimensional reconstruction module is used to add component labels to the second regional blocks in each of the regional block structures according to the preset proportion requirements of each component, so that each component in each layer of the regional block structure meets the preset proportion requirements, and obtain the three-dimensional gas diffusion layer; the three-dimensional gas diffusion layer includes a single-layer carbon fiber structure and a multi-layer regional block structure, each layer of the regional block structure includes a second regional block corresponding to the first regional block and a component of each second regional block, the second regional block is stacked on the corresponding three-dimensional first regional block, and the difference between the thickness of the second regional block and the thickness of the first regional block is less than a preset thickness threshold.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores at least one instruction or at least one program, and the at least one instruction or at least one program is loaded and executed by a processor to implement the three-dimensional image reconstruction method as described in any one of claims 1-5.

8. An electronic device, characterized in that: It includes at least one processor and a memory that is communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the at least one processor implements the image three-dimensional reconstruction method as described in any one of claims 1 to 5 by executing the instructions stored in the memory.

Images