Solid electrolyte interface analysis method, apparatus, device, and storage medium

Abstract

This invention belongs to the field of image analysis technology and discloses a method, apparatus, device, and storage medium for solid electrolyte interface analysis. The method includes: acquiring a solid electrolyte interface image; inputting the solid electrolyte interface image into a preset instance segmentation model for instance segmentation to obtain multiple lattice regions, classifying each lattice region to obtain the corresponding composition information; and completing solid electrolyte interface analysis based on the composition information corresponding to each lattice region. It should be noted that this solution distinguishes between lattice and non-lattice regions by instance segmentation of the solid electrolyte interface image, and then classifies the various partitions within the lattice regions to quickly obtain their composition. The composition of the lattice regions in the solid electrolyte interface can be directly determined through an end-to-end model, enabling batch image processing and improving analysis efficiency in video analysis.

Description

Technical Field

[0001] This invention relates to the field of image analysis technology, and in particular to a method, apparatus, device, and storage medium for solid electrolyte interface analysis. Background Technology

[0002] The SEI (solid electrolyte interphase) film is a crucial component of batteries, and analyzing its composition is of great significance for battery research. The SEI film is a thin layer structure formed between the electrode material and the electrolyte. It forms spontaneously during the first charge and discharge cycle. When the battery is first charged, the potential at the negative electrode surface drops to a certain level, causing a reduction reaction in the solvent molecules or lithium salts in the electrolyte. These reaction products deposit on the negative electrode surface, gradually accumulating to form a solid film layer covering the active material of the negative electrode. This film layer has a certain degree of conductivity, allowing lithium ions to pass through but blocking electrons, thus preventing further direct electrochemical reactions. The SEI film effectively isolates the electrode material and the electrolyte. The properties of the SEI film affect the lithium ion transport efficiency, thereby influencing battery performance indicators such as charge / discharge rate and capacity retention. Therefore, its formation and evolution process is crucial in battery material research.

[0003] Currently, electron microscopy is commonly used to characterize and analyze SEI, but this method suffers from significant issues in efficiency and accuracy.

[0004] The above content is only used to help understand the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention

[0005] The main objective of this invention is to provide a method, apparatus, device, and storage medium for solid electrolyte interface analysis, aiming to solve the technical problem of low efficiency in existing SEI membrane analysis.

[0006] To achieve the above objectives, the present invention provides a method for analyzing solid electrolyte interfaces, the method comprising the following steps:

[0007] Acquire a solid electrolyte interface image; input the solid electrolyte interface image into a preset instance segmentation model for instance segmentation to obtain multiple lattice regions, and classify each lattice region to obtain the composition information corresponding to each lattice region; complete the solid electrolyte interface analysis based on the composition information corresponding to each lattice region.

[0008] It should be noted that, compared to electron microscopy analysis, this method can quickly obtain the composition by segmenting and distinguishing lattice regions and non-lattice regions in the solid electrolyte interface image instance, and then classifying the various partitions in the lattice region. Since traditional electron microscopy analysis is inefficient and can only perform static analysis, this method can directly determine the composition of the lattice region in the solid electrolyte interface through an end-to-end model, batch process images, and apply it in video analysis. It can batch process each frame of image to obtain the evolution process of continuous lattice region images, thus improving the analysis efficiency.

[0009] In some embodiments, the preset instance segmentation model is trained in the following manner: acquiring image samples of a solid electrolyte interface, the image samples including interface images and lattice region annotation information corresponding to each interface image; inputting the interface images into an initial instance segmentation model to obtain prediction results for at least one lattice region; determining a loss value based on the prediction results of the lattice region and the corresponding lattice region annotation information, and adjusting the initial instance segmentation model based on the loss value until the prediction results converge to obtain the preset instance segmentation model.

[0010] It should be noted that a preset instance segmentation model can be obtained through a certain training process. By inputting the interface image into the initial instance segmentation model, prediction results of at least one lattice region can be obtained. The model can be continuously adjusted to obtain the preset instance segmentation model, thereby improving the analysis efficiency.

[0011] In some embodiments, before acquiring the image sample of the solid electrolyte interface, the method further includes: acquiring an original solid electrolyte interface image; inputting the original solid electrolyte interface image into a preset semantic segmentation model to determine an initial lattice region in the original solid electrolyte interface image; performing segmentation and boundary adjustment based on the initial lattice region to obtain multiple target lattice regions; inputting each target lattice region into a preset classification model to determine corresponding annotation information; and determining an image sample of the solid electrolyte interface based on the original solid electrolyte interface image and the annotation information corresponding to the target lattice regions.

[0012] Understandably, a high-performance model can be obtained through proper training. This solution uses two models to assist in annotation, enabling rapid model annotation while ensuring the efficiency of training sample generation.

[0013] In some embodiments, a lattice region to be analyzed is obtained; the interplanar spacing corresponding to the lattice region to be analyzed is determined through the lattice region to be analyzed; the composition information of the lattice region to be analyzed is determined by matching based on the interplanar spacing; and an initial classification model is trained based on the lattice region to be analyzed and the composition information of the lattice region to be analyzed to obtain a preset classification model.

[0014] It should be noted that the accuracy of the classification rules for the components is determined by how accurately they are defined. Therefore, this embodiment predicts the type by determining the interplanar spacing corresponding to the lattice region to be analyzed. This is because lattice spacing is relatively easy to calculate, has high computational efficiency, is suitable for batch operations, and has a good characterization effect on different components, thus improving the efficiency and convenience of sample label generation.

[0015] In some embodiments, the lattice region to be analyzed is analyzed to obtain a spectral image corresponding to the lattice region to be analyzed; multiple target spot features in the spectral image are identified; multiple spot spacings are determined based on the spot features; and the interplanar spacing corresponding to the lattice region to be analyzed is determined based on the multiple spot spacings.

[0016] Understandably, the calculation of lattice spacing involves identifying the features of some target spots in the lattice region. The distance between spots represents the distance in the lattice. Through certain conversions, the actual lattice spacing can be estimated, thus efficiently and quickly determining the lattice spacing.

[0017] In some embodiments, the crystal plane angles corresponding to the lattice spacing are determined based on the lattice region to be analyzed; the corresponding crystal plane indices are matched based on the lattice spacing and the corresponding crystal plane angles; the crystal plane indices are verified, and the composition information of the lattice region to be analyzed is determined based on the verification results.

[0018] It should be noted that matching composition information using lattice spacing may not be very accurate. Therefore, this embodiment proposes to determine the crystal plane angle, determine the composition information based on the crystal plane angle and lattice spacing, and verify it based on the crystal plane index corresponding to the composition information, thereby improving the accuracy of composition determination.

[0019] In some embodiments, classifying each of the lattice regions to obtain the composition information corresponding to each lattice region includes: determining the feature vector of each lattice region and inputting the feature vector into a classifier to determine the category score corresponding to the feature vector; determining the composition type of the lattice region and the probability information corresponding to the composition type based on the category score; and determining the composition information corresponding to the lattice region based on the composition type and the probability information corresponding to the composition type.

[0020] It should be noted that since the component information can be used in the video analysis process, in addition to the specific component content, this solution also provides the probability information of the components. The advantage of providing probability information is that by playing images from different frames continuously, we can see the changes in probability, which provides a reference for the formation, disappearance and changes of lattice regions, thus improving the analysis effect.

[0021] Secondly, to achieve the above objectives, the present invention also provides a solid electrolyte interface analysis device, the solid electrolyte interface analysis device comprising:

[0022] Thirdly, to achieve the above objectives, the present invention also provides a solid electrolyte interface analysis device, the solid electrolyte interface analysis device comprising: a memory, a processor, and a solid electrolyte interface analysis program stored in the memory and executable on the processor, the solid electrolyte interface analysis program being configured to implement the steps of the solid electrolyte interface analysis method as described above.

[0023] Fourthly, to achieve the above objectives, the present invention also provides a storage medium storing a solid electrolyte interface analysis program, wherein the solid electrolyte interface analysis program, when executed by a processor, implements the steps of the solid electrolyte interface analysis method as described above. Attached Figure Description

[0024] Figure 1 This is a schematic flowchart of the first embodiment of the solid electrolyte interface analysis method of the present invention;

[0025] Figure 2 This is a flowchart illustrating the second embodiment of the solid electrolyte interface analysis method of the present invention;

[0026] Figure 3 This is a schematic diagram of model prediction for an embodiment of the solid electrolyte interface analysis method of the present invention;

[0027] Figure 4 This is a schematic diagram comparing the original image and the frequency domain image of an embodiment of the solid electrolyte interface analysis method of the present invention;

[0028] Figure 5 This is a schematic diagram of the spot spacing in an embodiment of the solid electrolyte interface analysis method of the present invention;

[0029] Figure 6 This is a schematic diagram of the lattice composition database of an embodiment of the solid electrolyte interface analysis method of the present invention;

[0030] Figure 7 This is a verification diagram of an embodiment of the solid electrolyte interface analysis method of the present invention;

[0031] Figure 8 This is a structural block diagram of the first embodiment of the solid electrolyte interface analysis device of the present invention;

[0032] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0033] The embodiments of the technical solution of this application will now be described in detail with reference to the accompanying drawings. These embodiments are only used to more clearly illustrate the technical solution of this application and are therefore merely examples, and should not be used to limit the scope of protection of this application.

[0034] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0035] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0036] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0037] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0038] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0039] In the description of the embodiments of this application, the technical terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the embodiments of this application and simplifying the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the embodiments of this application.

[0040] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0041] The content disclosed in this application is mainly applied to the composition analysis of the SEI film. The SEI film, short for Solid Electrolyte Interface, is a crucial protective layer formed in lithium batteries. It typically forms spontaneously on the negative electrode surface during the first charge of the battery. The main components of the SEI film are various organic and inorganic compounds produced by the reduction reaction of solvents and lithium salts in the electrolyte on the surface of the negative electrode material. Since the composition of the SEI film directly affects the performance of the electrolyte in the battery, specifically, during the first charge and discharge of a liquid lithium-ion battery, the electrode material and the electrolyte react at the solid-liquid interface, forming a passivation layer covering the surface of the electrode material. This passivation layer is an interface layer with the characteristics of a solid electrolyte; it is an electronic insulator but an excellent conductor of Li+. Li+ can freely insert and extract through this passivation layer. When the battery is first charged, the potential on the negative electrode surface drops to a certain level, causing the solvent molecules or lithium salts in the electrolyte to undergo a reduction reaction. These reaction products are deposited on the negative electrode surface, and these decomposition products gradually accumulate, forming a solid film layer covering the negative electrode active material. This film layer possesses a certain degree of conductivity, allowing lithium ions to pass through but blocking electrons, thus preventing further direct electrochemical reactions. Therefore, this passivation film is called a "solid electrolyte interface" (SEI) film. In electron microscopy images of the SEI film, lattice regions are visible to the naked eye. These lattice regions represent specific material components. Identifying these lattice regions is crucial for SEI film analysis. The analysis of these lattice regions in the SEI film not only affects the basic electrochemical performance of the battery but also relates to its safety, reliability, and lifespan, providing important guidance for optimizing electrolytes and related materials.

[0042] In the context of SEI films, lattice regions refer to the ordered structures within the SEI film. These regions may be composed of inorganic compounds, which are generally more stable than amorphous portions. Ordered lattice structures are often more stable than amorphous structures, thus lattice regions contribute to improving the overall stability of the SEI film.

[0043] Therefore, this scheme proposes to automatically segment SEI membrane representation images by training an instance segmentation model, and calculate information such as the area of ​​the segmented regions using digital image processing methods, achieving an automated and efficient analysis method. This scheme proposes a training method for the SEI component instance segmentation model, which can effectively train an end-to-end machine learning model for SEI component analysis, assisting in SEI membrane research.

[0044] According to some embodiments of this application, such as Figure 1The diagram illustrates a method for analyzing the interface of solid electrolytes, the method comprising the following steps:

[0045] Acquire a solid electrolyte interface image; input the solid electrolyte interface image into a preset instance segmentation model for instance segmentation to obtain multiple lattice regions, and classify each lattice region to obtain the composition information corresponding to each lattice region; complete the solid electrolyte interface analysis based on the composition information corresponding to each lattice region.

[0046] It is understood that the solid electrolyte interface image is the image content of the solid electrolyte interface, which can be obtained through electron microscopy. In this embodiment, the acquired image is generally a video image. By decomposing the video image into several frames and inputting them into a preset instance segmentation model, a video image containing lattice region composition information is obtained to help technicians complete the analysis of the solid electrolyte interface. Specifically, the composition information may include the composition of the lattice region, the probability of the composition, or several possible compositions, etc.

[0047] Furthermore, for detailed analysis of the solid electrolyte interface (SEI), high-resolution images of the SEI can first be acquired using microscopic imaging techniques. These images are then input into a pre-trained instance segmentation model, which can identify and segment multiple lattice regions within the images. Further classification of each segmented lattice region determines the specific compositional information corresponding to each region. Finally, based on this compositional information, a comprehensive analysis of the SEI film is completed, including its structural features, compositional distribution, and potential functional properties. This contributes to a deeper understanding of the SEI film's role in batteries and provides a scientific basis for optimizing battery performance. Finally, reassembling the images into a video allows for clear observation of the evolution of the lattice regions, further enhancing the analytical results.

[0048] The specific prediction process of the model can be as follows: the solid electrolyte interface image is input into a preset instance segmentation model, and the image is segmented into fixed-size patches by limiting image feature extraction. These patches are then linearly mapped into vectors, and a feature representation rich in semantic information is generated through a multi-layer Transformer encoder. Next, multi-target recognition and classification need to be processed through an encoder-decoder structure. The encoder is responsible for extracting global context information from the image and introducing a set of learnable queries. Each query represents a potential target instance. These queries are the input of the decoder and are gradually refined by the decoder to finally generate the corresponding target box, which is equivalent to an instance of a lattice region. The composition information can then be obtained by predicting the category label corresponding to each query through a classifier, such as a fully connected layer with a softmax activation function.

[0049] It should be noted that, compared to electron microscopy analysis, this method can quickly obtain the composition by segmenting and distinguishing lattice regions and non-lattice regions in the solid electrolyte interface image instance, and then classifying the various partitions in the lattice region. Since traditional electron microscopy analysis is inefficient and can only perform static analysis, this method can directly determine the composition of the lattice region in the solid electrolyte interface through an end-to-end model, batch process images, and apply it in video analysis. It can batch process each frame of image to obtain the evolution process of continuous lattice region images, thus improving the analysis efficiency.

[0050] In some embodiments, such as Figure 2 As shown, the preset instance segmentation model is trained in the following manner: acquiring image samples of the solid electrolyte interface, the image samples including interface images and lattice region annotation information corresponding to each interface image; inputting the interface images into the initial instance segmentation model to obtain prediction results for at least one lattice region; determining a loss value based on the prediction results of the lattice region and the corresponding lattice region annotation information, and adjusting the initial instance segmentation model based on the loss value until the prediction results converge to obtain the preset instance segmentation model.

[0051] Understandably, the training process of a pre-defined instance segmentation model involves the following steps: First, high-quality image samples containing the solid electrolyte interface (SEI) are acquired. These images can be obtained through electron microscopy, capturing experimental SEI films, the SEI film formation process during generation, or through simulation, revealing the microstructure of the SEI. Each image is accompanied by detailed annotations, clearly indicating the location and boundaries of each lattice region in the image. Then, these annotated images are input into a pre-designed but not yet fully trained instance segmentation model. This model attempts to segment the input images to identify one or more lattice regions, then predicts each lattice region to obtain a high-dimensional vector representing the lattice composition type. The output is a preliminary prediction of the lattice region. Next, the model's prediction is compared with manually annotated real lattice region information. By calculating the difference between the two, a loss value is obtained, which quantifies the accuracy of the model's prediction. Based on this loss value, optimization algorithms (such as gradient descent) are used to adjust the model parameters, enabling the model to more accurately identify lattice regions in the next prediction. This process is not completed all at once, but rather through multiple iterations to gradually improve the model. After each iteration, the loss value is recalculated based on the new prediction results, and the model is further adjusted accordingly. This process continues until the model's prediction results no longer change significantly, i.e., convergence, or the preset performance index is reached. After multiple rounds of training and optimization, when the model's performance is stable and meets the expected requirements, a preset instance segmentation model is obtained. This end-to-end model can directly, automatically, and efficiently identify and label lattice regions from new solid electrolyte interface images without human intervention, providing strong support for materials science research.

[0052] It should be noted that a preset instance segmentation model can be obtained through a certain training process. By inputting the interface image into the initial instance segmentation model, prediction results of at least one lattice region can be obtained. The model can be continuously adjusted to obtain the preset instance segmentation model, thereby improving the analysis efficiency.

[0053] In some embodiments, before acquiring the image sample of the solid electrolyte interface, the method further includes: acquiring an original solid electrolyte interface image; inputting the original solid electrolyte interface image into a preset semantic segmentation model to determine an initial lattice region in the original solid electrolyte interface image; performing segmentation and boundary adjustment based on the initial lattice region to obtain multiple target lattice regions; inputting each target lattice region into a preset classification model to determine corresponding annotation information; and determining an image sample of the solid electrolyte interface based on the original solid electrolyte interface image and the annotation information corresponding to the target lattice regions.

[0054] It should be noted that a problem exists in the actual training process: how to obtain a large number of refined samples. This embodiment proposes to complete sample annotation by training two lightweight models. Specifically, for example, a lattice semantic segmentation model for SEI images is trained to segment lattice regions. For SEI electron microscope images, a batch of lattice region image data was manually annotated using a visual judgment method, and then a model was generated that is consistent with the original... Figure 1 A two-dimensional matrix of sample size is used, with lattice regions set to 1 and non-lattice regions set to 0. The labeled image and the processed 0+1 matrix are used as the model's input and output. Fine-tuning training is performed using a large image model, ultimately training a semantic segmentation model for lattices. A classification model is then trained for different lattice images to classify lattice images with different compositions. The predicted lattice region images are segmented into 256 or 512-sized images. Fourier transforms are then performed on these images, and information such as the lattice intercept is calculated in reciprocal space to infer the corresponding lattice composition. Data labeling of the corresponding composition in this way is performed, and the corresponding composition is encoded using one-hot encoding. The segmented lattice images and their corresponding one-hot encodings are used to train an image classification model, ultimately obtaining a lattice image classification model. Combining the above two models, the predicted lattice region images are segmented into smaller images. Classification prediction is performed on the segmented images, and the results are then mapped back to the original image and fine-tuned with annotations. The SEI image is segmented using a trained semantic segmentation model. The segmented lattice regions are then cut and fed into a classification model for prediction. Each predicted classification is mapped to a different color mask and then back onto the original image, resulting in instance segmentation masks for different lattice compositions. These masks are then manually corrected. An instance segmentation model for compositional regions is trained using fine-tuned labeled data. The obtained instance segmentation masks for different lattice compositions are used as the model's input and output to train an end-to-end compositional region instance segmentation model.

[0055] Understandably, a high-performance model can be obtained through proper training. This solution uses two models to assist in annotation, enabling rapid model annotation while ensuring the efficiency of training sample generation.

[0056] In some embodiments, a lattice region to be analyzed is obtained; the interplanar spacing corresponding to the lattice region to be analyzed is determined through the lattice region to be analyzed; the composition information of the lattice region to be analyzed is determined by matching based on the interplanar spacing; and an initial classification model is trained based on the lattice region to be analyzed and the composition information of the lattice region to be analyzed to obtain a preset classification model.

[0057] It is understandable that interplanar spacing refers to the perpendicular distance between two adjacent parallel crystal planes in a crystal structure. Inferring composition based primarily on information such as lattice spacing is a highly efficient method. The specific matching process involves matching the lattice spacing with a lattice database corresponding to each composition to determine the possible compositions present in the lattice regions. The inferred composition information is then used as the label information for the corresponding lattice regions. The initial classification model is trained using the lattice regions to be analyzed and their composition information, resulting in a mature and usable pre-defined classification model. This model serves as the basis for batch labeling image samples of solid electrolyte interfaces.

[0058] It should be noted that the accuracy of the classification rules for the components is determined by how accurately they are defined. Therefore, this embodiment predicts the type by determining the interplanar spacing corresponding to the lattice region to be analyzed. This is because lattice spacing is relatively easy to calculate, has high computational efficiency, is suitable for batch operations, and has a good characterization effect on different components, thus improving the efficiency and convenience of sample label generation.

[0059] In some embodiments, the lattice region to be analyzed is analyzed to obtain a spectral image corresponding to the lattice region to be analyzed; multiple target spot features in the spectral image are identified; multiple spot spacings are determined based on the spot features; and the interplanar spacing corresponding to the lattice region to be analyzed is determined based on the multiple spot spacings.

[0060] It should be noted that the prediction process for interplanar spacing can, for example, be achieved by obtaining the corresponding spectrum of the original image through Fourier transform, such as... Figure 4 The image shows the spectral distribution from the "lattice region to be analyzed" on the left to the right. It is evident that the spectral image on the right contains specks, and the characteristics of multiple target specks are... Figure 4 The spot image on the right side of the image can be used to determine the spot spacing by measuring and calculating the spacing between the spots. By statistically analyzing the spacing of multiple spots, the interplanar spacing of the lattice region to be analyzed can be predicted.

[0061] Specifically, such as Figure 5 As shown, there are 7 spots in the corresponding spectrum image. Select four of them, and then select a set of parallelograms in the spectrum image corresponding to the spots ( Figure 5 Perform length calibration of l1, l2, and l3 and corresponding measurement of the three included angles, and calculate the interplanar spacing d1, d2, and d3 according to the formula; taking d1 as an example, the formula is as follows:

[0062]

[0063] Where l0 is the line segment in the example figure, and the scale bar is the actual size corresponding to the line segment in the figure, i.e., 5 / nm. By conversion, the actual lengths of l1, l2, and l3, i.e., the spot spacing, can be obtained, and the interplanar spacings d1, d2, and d3 can be calculated.

[0064] Understandably, the calculation of lattice spacing involves identifying the features of some target spots in the lattice region. The distance between spots represents the distance in the lattice. Through certain conversions, the actual lattice spacing can be estimated, thus efficiently and quickly determining the lattice spacing.

[0065] In some embodiments, the crystal plane angles corresponding to the lattice spacing are determined based on the lattice region to be analyzed; the corresponding crystal plane indices are matched based on the lattice spacing and the corresponding crystal plane angles; the crystal plane indices are verified, and the composition information of the lattice region to be analyzed is determined based on the verification results.

[0066] Specifically, while determining the composition information of the lattice region to be analyzed solely based on the interplanar spacing is highly efficient, its accuracy may be limited. Therefore, this embodiment further proposes calculating the interplanar angle, where the interplanar angle refers to the angle between the normals of two different crystal planes in a crystal. Composition information is matched using the lattice spacing and the corresponding interplanar angle, such as... Figure 6 As shown, the closest composition is obtained by querying data based on the lattice spacing and the corresponding crystal plane angles.

[0067] Furthermore, to filter out invalid information, verification can also be performed based on crystal plane indices, such as... Figure 6 As shown, crystal plane indices can also be obtained by querying the lattice database corresponding to each composition. Crystal plane indices typically refer to Miller indices, a standard representation in crystallography used to identify different planes within a crystal. The predicted data is then compared with... Figure 6 The database data is compared, and three matching lattice indices R1, R2, and R3 are selected. If R1, R2, and R3 conform to the vector addition rule, then the component corresponding to this image can be identified as valid component information. Figure 7 As shown.

[0068] It should be noted that matching composition information using lattice spacing may not be very accurate. Therefore, this embodiment proposes to determine the crystal plane angle, determine the composition information based on the crystal plane angle and lattice spacing, and verify it based on the crystal plane index corresponding to the composition information, thereby improving the accuracy of composition determination.

[0069] In some embodiments, classifying each of the lattice regions to obtain the composition information corresponding to each lattice region includes: determining the feature vector of each lattice region and inputting the feature vector into a classifier to determine the category score corresponding to the feature vector; determining the composition type of the lattice region and the probability information corresponding to the composition type based on the category score; and determining the composition information corresponding to the lattice region based on the composition type and the probability information corresponding to the composition type.

[0070] It's important to note that after obtaining the preset instance segmentation model, for each predicted lattice region, its feature vector needs to be further extracted. These features can be based on image attributes such as shape, texture, color, and location, or they can be high-level features automatically learned from images using a convolutional neural network (CNN). The feature vector is a mathematical representation of the intrinsic properties within a lattice region, capturing the differences between different materials. Once the feature vectors of the lattice regions are obtained, they can be classified in the classifier part of the model. The classifier's task is to predict which compositional category each lattice region belongs to based on the feature vectors. Commonly used classifiers include Support Vector Machines (SVM), Random Forests, and Neural Networks. In this case, the classifier outputs a series of category scores for each lattice region, with each score corresponding to the probability that the lattice region might belong to a specific compositional type. For example, if there are three possible compositional types A, B, and C, the classifier will output three scores for each lattice region, representing the probability that it belongs to A, B, or C, respectively. The purpose of obtaining probability information is that, since most analyses are based on video image analysis, probability can be used to further analyze the evolution process of different lattice regions, for example... Figure 3 As shown.

[0071] It should be noted that since the component information can be used in the video analysis process, in addition to the specific component content, this solution also provides the probability information of the components. The advantage of providing probability information is that by playing images from different frames continuously, we can see the changes in probability, which provides a reference for the formation, disappearance and changes of lattice regions, thus improving the analysis effect.

[0072] Secondly, in order to achieve the above objectives, such as Figure 8 As shown, the present invention also provides a solid electrolyte interface analysis device, characterized in that the solid electrolyte interface analysis device includes: an acquisition module 10 for acquiring a solid electrolyte interface image; a processing module 20 for inputting the solid electrolyte interface image into a preset instance segmentation model for instance segmentation to obtain multiple lattice regions, and classifying each lattice region to obtain the composition information corresponding to each lattice region; the processing module 20 is used to complete the solid electrolyte interface analysis based on the composition information corresponding to each lattice region.

[0073] In some embodiments, the processing module 20 is used to acquire image samples of solid electrolyte interfaces, the image samples including interface images and lattice region annotation information corresponding to each interface image;

[0074] The interface image is input into the initial instance segmentation model to obtain the prediction result of at least one lattice region;

[0075] The loss value is determined based on the prediction results of the lattice region and the corresponding lattice region labeling information. The initial instance segmentation model is then adjusted based on the loss value until the prediction results converge, thus obtaining the preset instance segmentation model.

[0076] In some embodiments, the processing module 20 is used to acquire an original solid electrolyte interface image;

[0077] The original solid electrolyte interface image is input into a preset semantic segmentation model to determine the initial lattice region in the original solid electrolyte interface image;

[0078] Multiple target lattice regions are obtained by segmenting and boundary adjustment based on the initial lattice region;

[0079] Each target lattice region is input into a preset classification model to determine the corresponding annotation information;

[0080] Image samples of the solid electrolyte interface are determined based on the original solid electrolyte interface image and the annotation information corresponding to the target lattice region.

[0081] In some embodiments, the processing module 20 is used to acquire the lattice region to be analyzed;

[0082] The interplanar spacing of the lattice region to be analyzed is determined by the lattice region to be analyzed.

[0083] The composition information of the lattice region to be analyzed is determined by matching the interplanar spacing.

[0084] The initial classification model is trained based on the lattice region to be analyzed and its composition information to obtain the preset classification model.

[0085] In some embodiments, the processing module 20 is used to analyze the lattice region to be analyzed to obtain a spectral image corresponding to the lattice region to be analyzed;

[0086] Identify multiple target blob features in the spectral image;

[0087] The spacing between multiple spots is determined based on the spot characteristics;

[0088] The interplanar spacing corresponding to the lattice region to be analyzed is determined based on the spacing between the multiple spots.

[0089] In some embodiments, the processing module 20 is used to determine the crystal plane angle corresponding to the lattice spacing based on the lattice region to be analyzed;

[0090] The corresponding crystal plane index is matched according to the lattice spacing and the corresponding crystal plane angle;

[0091] The analysis is performed based on the crystal plane index, and the composition information of the lattice region to be analyzed is determined based on the analysis results.

[0092] In some embodiments, the processing module 20 is configured to determine the feature vector of each of the lattice regions and input the feature vector into a classifier to determine the category score corresponding to the feature vector;

[0093] The composition type of the lattice region and the probability information corresponding to the composition type are determined based on the category score.

[0094] The composition information corresponding to the lattice region is determined based on the composition type and the probability information corresponding to the composition type.

[0095] Thirdly, to achieve the above objectives, the present invention also provides a solid electrolyte interface analysis device, the solid electrolyte interface analysis device comprising: a memory, a processor, and a solid electrolyte interface analysis program stored in the memory and executable on the processor, the solid electrolyte interface analysis program being configured to implement the steps of the solid electrolyte interface analysis method as described above.

[0096] Fourthly, to achieve the above objectives, the present invention also provides a storage medium storing a solid electrolyte interface analysis program, wherein the solid electrolyte interface analysis program, when executed by a processor, implements the steps of the solid electrolyte interface analysis method as described above.

[0097] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and not to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. These modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application, and they should all be covered within the scope of the claims and specification of this application. In particular, as long as there is no structural conflict, the various technical features mentioned in the embodiments can be combined in any way. This application is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. A method for analyzing the interface of solid electrolytes, characterized in that, The solid electrolyte interface analysis method includes: Acquire solid electrolyte interface images; The solid electrolyte interface image is input into a preset instance segmentation model for instance segmentation to obtain multiple lattice regions. Each lattice region is then classified to obtain the composition information corresponding to each lattice region. Solid electrolyte interface analysis is performed based on the composition information corresponding to each lattice region. The composition information corresponding to the lattice region includes at least one of the composition of the lattice region and the probability of the composition; The step of performing solid electrolyte interface analysis based on the composition information corresponding to each lattice region includes: The composition information of the lattice region corresponding to each frame of solid electrolyte interface image is stitched together to obtain the composition information corresponding to the continuous solid electrolyte image; The component information corresponding to the continuous images of solid electrolytes is used to complete the interface analysis of solid electrolytes. The pre-defined classification model is trained in the following ways: Obtain the lattice region to be analyzed; The interplanar spacing of the lattice region to be analyzed is determined by the lattice region to be analyzed. The composition information of the lattice region to be analyzed is determined by matching the interplanar spacing. The initial classification model is trained based on the lattice region to be analyzed and its composition information to obtain the preset classification model.

2. The method as described in claim 1, characterized in that, The preset instance segmentation model is trained in the following ways: Acquire image samples of solid electrolyte interfaces, the image samples including interface images and lattice region annotation information corresponding to each interface image; The interface image is input into the initial instance segmentation model to obtain the prediction result of at least one lattice region; The loss value is determined based on the prediction results of the lattice region and the corresponding lattice region labeling information. The initial instance segmentation model is then adjusted based on the loss value until the prediction results converge, thus obtaining the preset instance segmentation model.

3. The method as described in claim 2, characterized in that, Before acquiring the image sample of the solid electrolyte interface, the method further includes: Obtain the original solid electrolyte interface image; The original solid electrolyte interface image is input into a preset semantic segmentation model to determine the initial lattice region in the original solid electrolyte interface image; Multiple target lattice regions are obtained by segmenting and boundary adjustment based on the initial lattice region; Each target lattice region is input into a preset classification model to determine the corresponding annotation information; Image samples of the solid electrolyte interface are determined based on the original solid electrolyte interface image and the annotation information corresponding to the target lattice region.

4. The method as described in claim 1, characterized in that, The step of determining the interplanar spacing corresponding to the lattice region to be analyzed through the lattice region to be analyzed includes: The spectral image corresponding to the lattice region to be analyzed is obtained by analyzing the lattice region to be analyzed; Identify multiple target blob features in the spectral image; The spacing between multiple spots is determined based on the spot characteristics; The interplanar spacing corresponding to the lattice region to be analyzed is determined based on the spacing between the multiple spots.

5. The method as described in claim 1, characterized in that, The step of determining the composition information of the lattice region to be analyzed by matching based on the interplanar spacing includes: Determine the crystal plane angles corresponding to the lattice spacing based on the lattice region to be analyzed; The corresponding crystal plane index is matched according to the lattice spacing and the corresponding crystal plane angle; The analysis is performed based on the crystal plane index, and the composition information of the lattice region to be analyzed is determined based on the analysis results.

6. The method as described in claim 1, characterized in that, The step of classifying each of the lattice regions to obtain the composition information corresponding to each of the lattice regions includes: The feature vectors of each lattice region are determined, and the feature vectors are input into a classifier to determine the category score corresponding to the feature vectors; The composition type of the lattice region and the probability information corresponding to the composition type are determined based on the category score. The composition information corresponding to the lattice region is determined based on the composition type and the probability information corresponding to the composition type.

7. A solid electrolyte interface analysis device, characterized in that, The solid electrolyte interface analysis device includes: The acquisition module is used to acquire images of the solid electrolyte interface. The processing module is used to input the solid electrolyte interface image into a preset instance segmentation model for instance segmentation to obtain multiple lattice regions, and classify each lattice region to obtain the composition information corresponding to each lattice region. The processing module is used to perform solid electrolyte interface analysis based on the composition information corresponding to each lattice region. The composition information corresponding to the lattice region includes at least one of the composition of the lattice region and the probability of the composition; The step of performing solid electrolyte interface analysis based on the composition information corresponding to each lattice region includes: The composition information of the lattice region corresponding to each frame of solid electrolyte interface image is stitched together to obtain the composition information corresponding to the continuous solid electrolyte image; The component information corresponding to the continuous images of solid electrolytes is used to complete the interface analysis of solid electrolytes. The processing module is used to acquire the lattice region to be analyzed; determine the interplanar spacing corresponding to the lattice region to be analyzed; determine the composition information of the lattice region to be analyzed by matching the interplanar spacing; and train an initial classification model based on the lattice region to be analyzed and its composition information to obtain a preset classification model.

8. A solid electrolyte interface analysis device, characterized in that, The device includes: a memory, a processor, and a solid electrolyte interface analysis program stored in the memory and executable on the processor, the solid electrolyte interface analysis program being configured to implement the steps of the solid electrolyte interface analysis method as described in any one of claims 1 to 6.

9. A storage medium, characterized in that, The storage medium stores a solid electrolyte interface analysis program, which, when executed by a processor, implements the steps of the solid electrolyte interface analysis method as described in any one of claims 1 to 6.

Images