Aberration parameter determination method and device of transmission electron microscope, equipment and storage medium

Abstract

The invention discloses an aberration parameter determination method and device of a transmission electron microscope, equipment and a storage medium, and relates to the field of transmission electron microscopes, and the method comprises the steps: obtaining an original training data set, and obtaining an original to-be-measured data set through a target transmission electron microscope; processing a first Sohn ring picture in the original training data set to obtain a binary training picture, and processing a second Sohn ring picture in the original to-be-measured data set to obtain a binary target picture; constructing a target training set and a target verification set based on the binarized training pictures and the corresponding simulation labels to train a target parameter determination model; inputting the binarized target picture into a target parameter determination model to determine a target parameter, and optimizing an aberration corrector of the target transmission electron microscope based on the target parameter; wherein the target parameter is a second contrast transfer function parameter corresponding to the second Sohn ring picture. According to the invention, the aberration information contained in the Sohn ring corresponding to the transmission electron microscope can be accurately detected.

Description

Technical Field

[0001] The present invention relates to the field of transmission electron microscopes, and in particular to a method, device, equipment and storage medium for determining aberration parameters of a transmission electron microscope. Background Art

[0002] In transmission electron microscopy, multiple aberration parameters affect imaging accuracy, including spherical aberration, underfocus, second-order astigmatism, higher-order astigmatism, and coma. Third-order spherical aberration, underfocus, and second-order astigmatism are the most significant factors affecting the image. Existing methods for detecting aberrations in Thorne rings (also known as amorphous rings) mostly use elliptical equations to approximate the ring's one-dimensional power spectrum, then solve the equation to obtain the individual aberration values. These methods introduce errors in the process of obtaining the one-dimensional power spectrum using approximate methods, and because they require manual judgment regarding which equation (elliptical, parabolic, or hyperbolic) to use for fitting, they cannot automatically detect aberrations.

[0003] In summary, how to accurately detect the aberration information contained in the Thorne ring corresponding to the transmission electron microscope is a problem that needs to be solved urgently. Summary of the Invention

[0004] In view of this, the present invention aims to provide a method, apparatus, device, and storage medium for determining aberration parameters of a transmission electron microscope, which can accurately detect the aberration information contained in the Thorne ring corresponding to the transmission electron microscope. The specific scheme is as follows:

[0005] In a first aspect, the present application discloses a method for determining aberration parameters of a transmission electron microscope, comprising:

[0006] Acquire an original training data set, and acquire an original data set to be measured using a target transmission electron microscope; the original training data set includes first Thorne ring images and first contrast transfer function parameters corresponding to each of the first Thorne ring images; the first contrast transfer function parameters include third-order spherical aberration, defocus, angle of second-order astigmatism, and magnitude of second-order astigmatism; the original data set to be measured includes second Thorne ring images for which contrast transfer function parameters are to be determined;

[0007] Processing all the first Thorne ring images in the original training data set to obtain binary training images corresponding to the first Thorne ring images, and processing all the second Thorne ring images in the original to-be-measured data set to obtain binary target images corresponding to the second Thorne ring images;

[0008] A target training set and a target validation set are constructed based on the binarized training images and the corresponding simulated labels, and a target parameter determination model is trained using the target training set and the target validation set; the simulated labels are the first contrast transfer function parameters corresponding to the binarized training images, and the target parameter determination model is a convolutional neural network model for determining contrast transfer function parameters corresponding to Thorne ring images;

[0009] The binarized target image is input into the target parameter determination model to determine target parameters, so as to optimize the aberration corrector of the target transmission electron microscope based on the target parameters; wherein the target parameters are second contrast transfer function parameters corresponding to the second Thorne ring image.

[0010] Optionally, obtaining the original training data set includes:

[0011] A preset number of first Thorne ring images are acquired using a preset transmission electron microscope with known contrast transfer function parameters, and an original training data set is constructed using the first Thorne ring images and the first contrast transfer function parameters corresponding to each of the first Thorne ring images.

[0012] Optionally, obtaining the original training data set includes:

[0013] determining a first contrast transfer function parameter;

[0014] Simulating an imaging process of a transmission electron microscope using the first contrast transfer function parameters to obtain a first Thorne ring image corresponding to the first contrast transfer function parameters;

[0015] The original training data set is constructed using the first Thorne ring image and the corresponding first contrast transfer function parameters.

[0016] Optionally, simulating the imaging process of a transmission electron microscope using the first contrast transfer function parameters to obtain a first Thorne ring image corresponding to the first contrast transfer function parameters includes:

[0017] determining an aberration function corresponding to the first contrast transfer function parameter using the aberration parameter of the first transmission electron microscope;

[0018] determining a one-dimensional contrast transfer function corresponding to the aberration function based on the aberration function;

[0019] The one-dimensional contrast transfer function is expanded in two dimensions to obtain a first Thorne ring image corresponding to the first contrast transfer function parameters.

[0020] Optionally, processing all the first Thorne ring images in the original training data set to obtain binary training images corresponding to the first Thorne ring images, and processing all the second Thorne ring images in the original to-be-measured data set to obtain binary target images corresponding to the second Thorne ring images, includes:

[0021] Performing grayscale processing and noise reduction processing on all the first Thorne ring images in the original training data set to obtain first noise-reduced images;

[0022] performing thresholding processing on the first denoised image based on a first preset threshold to obtain a binary training image corresponding to the first Thorne ring image;

[0023] Performing grayscale processing and noise reduction processing on all the second Thorne ring images in the original data set to be measured to obtain second noise-reduced images;

[0024] The second denoised image is thresholded based on a second preset threshold to obtain a binarized target image corresponding to the second Thorne ring image.

[0025] Optionally, constructing a target training set and a target validation set based on the binarized training images and corresponding simulated labels, and using the target training set and the target validation set to train a target parameter determination model includes:

[0026] Constructing a target training set and a target verification set based on a preset ratio using the binarized training images and corresponding simulated labels;

[0027] Using the target training set, an initial parameter determination model is trained based on preset training parameters to obtain a trained parameter determination model; the preset training parameters include a learning rate, a batch size, and a number of training rounds; and the initial parameter determination model is a convolutional neural network model;

[0028] After each iteration is completed, the model parameters of the current post-training parameter determination model are recorded, and the target verification set is input into the current post-training parameter determination model to obtain an error convergence curve;

[0029] If the current number of iteration rounds is equal to the number of training rounds, target model parameters are determined from all the model parameters based on the error convergence curve, and the trained parameter determination model corresponding to the target model parameters is determined as the target parameter determination model.

[0030] Optionally, the target parameter determines that the activation function of the last fully connected layer of the model is a linear activation function.

[0031] In a second aspect, the present application discloses a device for determining aberration parameters of a transmission electron microscope, comprising:

[0032] a data set acquisition module, configured to acquire an original training data set and to acquire an original data set to be measured using a target transmission electron microscope; the original training data set includes first Thorne ring images and first contrast transfer function parameters corresponding to each of the first Thorne ring images; the first contrast transfer function parameters include third-order spherical aberration, defocus, angle of second-order astigmatism, and magnitude of second-order astigmatism; the original data set to be measured includes second Thorne ring images for which contrast transfer function parameters are to be determined;

[0033] an image processing module, configured to process all the first Thorne's ring images in the original training data set to obtain binary training images corresponding to the first Thorne's ring images, and to process all the second Thorne's ring images in the original to-be-measured data set to obtain binary target images corresponding to the second Thorne's ring images;

[0034] a model training module, configured to construct a target training set and a target validation set based on the binarized training images and corresponding simulated labels, and to train a target parameter determination model using the target training set and the target validation set; wherein the simulated labels are the first contrast transfer function parameters corresponding to the binarized training images, and the target parameter determination model is a convolutional neural network model for determining contrast transfer function parameters corresponding to Thorne ring images;

[0035] A parameter determination module is used to input the binarized target image into the target parameter determination model to determine target parameters, so as to optimize the aberration corrector of the target transmission electron microscope based on the target parameters; wherein the target parameters are second contrast transfer function parameters corresponding to the second Thorne ring image.

[0036] In a third aspect, the present application discloses an electronic device, comprising:

[0037] Memory, used to store computer programs;

[0038] The processor is used to execute the computer program to implement the aforementioned method for determining the aberration parameters of the transmission electron microscope.

[0039] In a fourth aspect, the present application discloses a computer-readable storage medium for storing a computer program, wherein the computer program, when executed by a processor, implements the aforementioned method for determining aberration parameters of a transmission electron microscope.

[0040] In the present application, when determining the aberration parameters of a transmission electron microscope, an original training data set is obtained, and an original data set to be measured is obtained using a target transmission electron microscope; the original training data set includes a first Thorne ring image and a first contrast transfer function parameter corresponding to each of the first Thorne ring images; the first contrast transfer function parameter includes a third-order spherical aberration, a defocus amount, an angle of second-order astigmatism, and a magnitude of second-order astigmatism; the original data set to be measured includes a second Thorne ring image whose contrast transfer function parameters are to be determined; all the first Thorne ring images in the original training data set are processed to obtain a binarized training image corresponding to the first Thorne ring image, and all the second Thorne ring images in the original data set to be measured are processed to obtain the A binarized target image corresponding to the second Thorne ring image; a target training set and a target verification set are constructed based on the binarized training image and the corresponding simulation labels, and a target parameter determination model is trained using the target training set and the target verification set; the simulation labels are the first contrast transfer function parameters corresponding to the binarized training image, and the target parameter determination model is a convolutional neural network model for determining the contrast transfer function parameters corresponding to the Thorne ring image; the binarized target image is input into the target parameter determination model to determine the target parameters, so as to optimize the aberration corrector of the target transmission electron microscope based on the target parameters; wherein the target parameters are the second contrast transfer function parameters corresponding to the second Thorne ring image. It can be seen that the present application first obtains a first Thorne ring image with known contrast transfer function parameters, and after binarizing the first Thorne ring image, uses the obtained binarized training image and the corresponding simulation label (that is, the contrast transfer function parameters corresponding to the first Thorne ring image) to construct a target training set and a target verification set, so that the convolutional neural network model (that is, the target parameter determination model) trained using the target training set and the target verification set can identify the aberration information contained in the binarized image corresponding to the newly input Thorne ring image (that is, the second Thorne ring image), thereby quickly determining the target parameters of the target transmission electron microscope that took the Thorne ring image, and realizing accurate detection of the aberration information contained in the Thorne ring corresponding to the transmission electron microscope. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0042] Figure 1 This is a flow chart of a method for determining aberration parameters of a transmission electron microscope disclosed in this application;

[0043] Figure 2 This is a schematic diagram of the first Thorne ring disclosed in this application;

[0044] Figure 3 This is a schematic diagram of a specific target parameter determination model structure disclosed in this application;

[0045] Figure 4 This is a specific target parameter correctness verification diagram disclosed in this application, where Figure 4 (a) and Figure 4 (c) is a picture of the second Thorne ring collected using a transmission electron microscope with different objectives. Figure 4 (b) For the use of Figure 4 (a) The Thorne ring obtained by aberration simulation of the corresponding target parameters and Figure 4 (a) The result image after splicing, Figure 4 (d) For the use of Figure 4 (c) The Thorne ring obtained by aberration simulation of the corresponding target parameters and Figure 4 (c) The resulting image after splicing;

[0046] Figure 5 This is a schematic structural diagram of a device for determining aberration parameters of a transmission electron microscope disclosed in this application;

[0047] Figure 6 This is a structural diagram of an electronic device disclosed in this application. DETAILED DESCRIPTION

[0048] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 efforts are within the scope of protection of the present invention.

[0049] Existing methods for detecting aberrations from Thorne rings (also known as amorphous rings) mostly use elliptical equations or other methods to approximate the one-dimensional power spectrum of the amorphous rings, and then use the idea of ​​solving equations to obtain the various aberration values. Such methods will introduce errors in the process of obtaining the one-dimensional power spectrum using approximate methods. Moreover, because these methods require manual judgment of which equation (elliptical equation, parabolic equation, or hyperbolic equation) to use for fitting, they cannot automatically detect aberrations. To address the above technical problems, the present application discloses a method for determining aberration parameters of a transmission electron microscope, which can accurately detect the aberration information contained in the Thorne rings corresponding to the transmission electron microscope.

[0050] See also Figure 1As shown, an embodiment of the present invention discloses a method for determining aberration parameters of a transmission electron microscope, comprising:

[0051] Step S11: Acquire an original training data set, and acquire an original data set to be measured using a target transmission electron microscope; the original training data set includes first Thorne ring images and first contrast transfer function parameters corresponding to each of the first Thorne ring images; the first contrast transfer function parameters include third-order spherical aberration, defocus, angle of second-order astigmatism, and magnitude of second-order astigmatism; the original data set to be measured includes second Thorne ring images whose contrast transfer function parameters are to be determined.

[0052] In this embodiment, it is necessary to first obtain an original training data set that can be used to train the model, and then use the target transmission electron microscope whose aberration parameters are to be determined to obtain the original data set to be measured. The original training data set includes the first Thorne ring image and the first contrast transfer function parameters corresponding to each first Thorne ring image; the first contrast transfer function parameters include the third-order spherical aberration, the amount of defocus, the angle of the second-order astigmatism, and the size of the second-order astigmatism; accordingly, the original data set to be measured includes the second Thorne ring image whose contrast transfer function parameters are to be determined. Figure 2 Shown is a portion of the first Thorne ring image that can be used to train the target parameter determination model.

[0053] In one specific embodiment, to obtain an original training dataset, a preset number of first Thorne ring images can be acquired using a preset transmission electron microscope (TEM) with known contrast transfer function parameters. The original training dataset is then constructed using the first Thorne ring images and the first contrast transfer function parameters corresponding to each first Thorne ring image. In another specific embodiment, the first contrast transfer function parameters can be first determined; then, the first contrast transfer function parameters can be used to simulate the imaging process of the transmission electron microscope to obtain first Thorne ring images corresponding to the first contrast transfer function parameters. Finally, the original training dataset is constructed using the first Thorne ring images and the corresponding first contrast transfer function parameters.

[0054] In this embodiment, when the aperture function and the attenuation envelope function are ignored, the contrast transfer function It can be expressed as:

[0055] ;

[0056] When only considering the third-order spherical aberration, defocus and second-order astigmatism, the aberration function It can be expressed as:

[0057] ;

[0058] in, represents the coordinate vector in Fourier space, represents the electron wavelength, Indicates the amount of defocus. It is the third-order spherical aberration. describes the azimuthal variation of the defocus amount, where represents the amplitude of the second-order astigmatism and is always positive, while is the angle of second-order astigmatism. is the azimuth of each point on the ring. If combined with the one-dimensional power spectrum With known spherical aberration, defocus and second-order astigmatism, the following formula can be used to solve:

[0059] ;

[0060] Therefore, in this embodiment, if the original training dataset is obtained by simulating the imaging process of a transmission electron microscope, the imaging process of the transmission electron microscope is simulated using the first contrast transfer function parameters to obtain a first Thorne ring image corresponding to the first contrast transfer function parameters. Specifically, this may include: determining an aberration function corresponding to the first contrast transfer function parameters using aberration parameters of the first transmission electron microscope; determining a one-dimensional contrast transfer function corresponding to the aberration function based on the aberration function; and performing a two-dimensional expansion of the one-dimensional contrast transfer function to obtain the first Thorne ring image corresponding to the first contrast transfer function parameters. In practical applications, the effects of the aperture function and the attenuation envelope function on the simulation process may also be considered as needed during the process of simulating the imaging process of a transmission electron microscope to obtain the original training dataset.

[0061] Step S12: Process all the first Thorne ring images in the original training data set to obtain binary training images corresponding to the first Thorne ring images, and process all the second Thorne ring images in the original data set to be measured to obtain binary target images corresponding to the second Thorne ring images.

[0062] In this embodiment, the grayscale values ​​of the acquired first Thorne ring images are distributed between 0 and 255. To highlight the ring features and separate them from the background, the first Thorne ring images need to be binarized before the target parameter determination model is trained using the original training dataset. Accordingly, the second Thorne ring images also need to be binarized before the target parameters corresponding to the second Thorne ring images are measured. Specifically, when binarizing the first Thorne ring images, grayscale processing and noise reduction processing can be performed on all first Thorne ring images in the original training dataset to obtain a first denoised image. The first denoised image can then be thresholded based on a first preset threshold to obtain a binary training image corresponding to the first Thorne ring image. When binarizing the second Thorne ring images, grayscale processing and noise reduction processing can be performed on all second Thorne ring images in the original dataset to be measured to obtain a second denoised image. The second denoised image can then be thresholded based on a second preset threshold to obtain a binary target image corresponding to the second Thorne ring image. In a specific implementation, when performing the binarization process, an adaptive threshold algorithm or other algorithms may be used to perform the binarization process on the acquired Thorne ring image.

[0063] Step S13: constructing a target training set and a target verification set based on the binarized training image and the corresponding simulated label, and using the target training set and the target verification set to train a target parameter determination model; the simulated label is the first contrast transfer function parameter corresponding to the binarized training image, and the target parameter determination model is a convolutional neural network model for determining the contrast transfer function parameter corresponding to the Thorne ring image.

[0064] In this embodiment, the convolutional neural network model can simultaneously detect aberration parameters such as third-order spherical aberration, defocus, and the size and angle of second-order astigmatism from a two-dimensional power spectrum by identifying the position and features of the transformation. As a supervised learning method, the convolutional neural network model requires a large number of images and their corresponding labels for training. In a specific embodiment, a target training set and a target validation set are constructed based on the binarized training images and the corresponding simulated labels, and the target training set and the target validation set are used to train a target parameter determination model, including: constructing a target training set and a target validation set based on a preset ratio using the binarized training images and the corresponding simulated labels; using the target training set to train an initial parameter determination model based on preset training parameters to obtain a post-training parameter determination model; the preset training parameters include a learning rate, a batch size, and a number of training rounds; the initial parameter determination model is a convolutional neural network model; after each iteration, the model parameters of the current post-training parameter determination model are recorded, and the target validation set is input into the current post-training parameter determination model to obtain an error convergence curve; if the current number of iteration rounds is equal to the number of training rounds, the target model parameters are determined from all model parameters based on the error convergence curve, and the post-training parameter determination model corresponding to the target model parameters is determined as the target parameter determination model. The target parameter determines that the activation function of the last fully connected layer of the model is a linear activation function. Figure 3 What is shown is a structural diagram of a specific target parameter determination model.

[0065] Step S14: inputting the binarized target image into the target parameter determination model to determine target parameters, so as to optimize the aberration corrector of the target transmission electron microscope based on the target parameters; wherein the target parameters are second contrast transfer function parameters corresponding to the second Thorne ring image.

[0066] In this embodiment, after obtaining the target parameter determination model, the binarized target image can be input into the target parameter determination model, so as to use the target parameter determination model to quickly and accurately determine the target parameters corresponding to the binarized target image, that is, the second contrast transfer function parameters corresponding to the second Thorne ring image corresponding to the binarized target image, and then the target parameters can be used to optimize the aberration corrector of the target transmission electron microscope corresponding to the second Thorne ring image. It can be understood that the target parameters can also be used for wave function reconstruction and image simulation. Figure 4The figure shows a schematic diagram of a specific target parameter correctness verification disclosed in this application, wherein Figures (a) and (c) are second Thorne ring images collected using different target transmission electron microscopes, Figure (b) is a result image obtained by splicing the Thorne ring obtained by using the target parameter determination model to obtain the target parameters corresponding to Figure (a) (the second row in Table 1) and performing aberration simulation based on the target parameters, and Figure (a), and Figure (d) is a result image obtained by splicing the Thorne ring obtained by using the target parameter determination model to obtain the target parameters corresponding to Figure (c) (the third row in Table 1) and performing aberration simulation based on the target parameters, and Figure (c). It can be seen that the rings in Figures (b) and (d) are well matched, which also proves the accuracy and practicality of the aberration parameter determination method for a transmission electron microscope provided in this embodiment.

[0067] Table 1

[0068]

[0069] It can be seen that the present application first obtains a first Thorne ring image with known contrast transfer function parameters, and after binarizing the first Thorne ring image, uses the obtained binarized training image and the corresponding simulation label (that is, the contrast transfer function parameters corresponding to the first Thorne ring image) to construct a target training set and a target verification set, so that the convolutional neural network model (that is, the target parameter determination model) trained using the target training set and the target verification set can identify the aberration information contained in the binarized image corresponding to the newly input Thorne ring image (that is, the second Thorne ring image), thereby quickly determining the target parameters of the target transmission electron microscope that took the Thorne ring image, and realizing accurate detection of the aberration information contained in the Thorne ring corresponding to the transmission electron microscope.

[0070] See also Figure 5 As shown, the present application discloses a device for determining aberration parameters of a transmission electron microscope, comprising:

[0071] The data set acquisition module 11 is configured to acquire an original training data set and to acquire an original data set to be measured using a target transmission electron microscope; the original training data set includes first Thorne ring images and first contrast transfer function parameters corresponding to each of the first Thorne ring images; the first contrast transfer function parameters include third-order spherical aberration, defocus, angle of second-order astigmatism, and magnitude of second-order astigmatism; the original data set to be measured includes second Thorne ring images for which contrast transfer function parameters are to be determined;

[0072] An image processing module 12 is configured to process all the first Thorne's ring images in the original training data set to obtain binary training images corresponding to the first Thorne's ring images, and to process all the second Thorne's ring images in the original to-be-measured data set to obtain binary target images corresponding to the second Thorne's ring images;

[0073] A model training module 13 is configured to construct a target training set and a target validation set based on the binarized training images and the corresponding simulated labels, and to train a target parameter determination model using the target training set and the target validation set; wherein the simulated labels are the first contrast transfer function parameters corresponding to the binarized training images, and the target parameter determination model is a convolutional neural network model for determining contrast transfer function parameters corresponding to Thorne ring images;

[0074] The parameter determination module 14 is used to input the binarized target image into the target parameter determination model to determine target parameters, so as to optimize the aberration corrector of the target transmission electron microscope based on the target parameters; wherein the target parameters are second contrast transfer function parameters corresponding to the second Thorne ring image.

[0075] It can be seen that the present application first obtains a first Thorne ring image with known contrast transfer function parameters, and after binarizing the first Thorne ring image, uses the obtained binarized training image and the corresponding simulation label (that is, the contrast transfer function parameters corresponding to the first Thorne ring image) to construct a target training set and a target verification set, so that the convolutional neural network model (that is, the target parameter determination model) trained using the target training set and the target verification set can identify the aberration information contained in the binarized image corresponding to the newly input Thorne ring image (that is, the second Thorne ring image), thereby quickly determining the target parameters of the target transmission electron microscope that took the Thorne ring image, and realizing accurate detection of the aberration information contained in the Thorne ring corresponding to the transmission electron microscope.

[0076] In a specific implementation, the data set acquisition module 11 may specifically include:

[0077] The first data set acquisition submodule is used to acquire a preset number of first Thorne ring images using a preset transmission electron microscope with known contrast transfer function parameters, and to construct an original training data set using the first Thorne ring images and the first contrast transfer function parameters corresponding to each of the first Thorne ring images.

[0078] In a specific implementation, the data set acquisition module 11 may specifically include:

[0079] A first parameter determination submodule, configured to determine a first contrast transfer function parameter;

[0080] An image acquisition submodule, configured to simulate an imaging process of a transmission electron microscope using the first contrast transfer function parameters to obtain a first Thorne ring image corresponding to the first contrast transfer function parameters;

[0081] The data set acquisition submodule is configured to construct the original training data set using the first Thorne ring image and the corresponding first contrast transfer function parameters.

[0082] In a specific implementation, the first image acquisition submodule may specifically include:

[0083] an aberration function determining unit, configured to determine an aberration function corresponding to the first contrast transfer function parameters using the aberration parameters of the first transmission electron microscope;

[0084] a one-dimensional contrast transfer function determining unit, configured to determine a one-dimensional contrast transfer function corresponding to the aberration function based on the aberration function;

[0085] The Thorne ring image acquisition unit is configured to perform a two-dimensional expansion on the one-dimensional contrast transfer function to obtain a first Thorne ring image corresponding to the first contrast transfer function parameters.

[0086] In a specific implementation, the image processing module 12 may include:

[0087] A first image denoising submodule, configured to perform grayscale processing and denoise processing on all the first Thorne ring images in the original training data set to obtain a first denoised image;

[0088] A first image binarization submodule, configured to perform thresholding processing on the first denoised image based on a first preset threshold value to obtain a binarized training image corresponding to the first Thorne ring image;

[0089] A second image denoising submodule, configured to perform grayscale processing and denoise processing on all the second Thorne ring images in the original data set to be measured to obtain a second denoised image;

[0090] The second image binarization submodule is configured to perform thresholding processing on the second denoised image based on a second preset threshold value to obtain a binarized target image corresponding to the second Thorne ring image.

[0091] In a specific implementation, the model training module 13 may specifically include:

[0092] A data set construction submodule is used to construct a target training set and a target verification set based on a preset ratio using the binarized training images and corresponding simulated labels;

[0093] a model training submodule, configured to train an initial parameter determination model based on preset training parameters using the target training set to obtain a trained parameter determination model; the preset training parameters include a learning rate, a batch size, and a number of training rounds; and the initial parameter determination model is a convolutional neural network model;

[0094] A model verification module is used to record the model parameters of the current post-training parameter determination model after each iteration is completed, and input the target verification set into the current post-training parameter determination model to obtain an error convergence curve;

[0095] The model determination submodule is used to determine the target model parameters from all the model parameters based on the error convergence curve if the current number of iteration rounds is equal to the number of training rounds, and determine the trained parameter determination model corresponding to the target model parameters as the target parameter determination model.

[0096] Furthermore, the embodiment of the present application also discloses an electronic device, Figure 6 This is a structural diagram of an electronic device 20 according to an exemplary embodiment. The content in the diagram should not be considered as any limitation to the scope of application of the present application.

[0097] Figure 6 This is a schematic diagram of the structure of an electronic device 20 provided in an embodiment of the present application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 is used to store a computer program, which is loaded and executed by the processor 21 to implement the relevant steps of the method for determining aberration parameters of a transmission electron microscope disclosed in any of the aforementioned embodiments. Furthermore, the electronic device 20 in this embodiment may specifically be an electronic computer.

[0098] In this embodiment, the power supply 23 is used to provide operating voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and the external device. The communication protocol it follows is any communication protocol that can be applied to the technical solution of this application and is not specifically limited here; the input and output interface 25 is used to obtain external input data or output data to the outside world. Its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0099] In addition, the memory 22 as a carrier for resource storage can be a read-only memory, random access memory, disk or optical disk, etc. The resources stored thereon can include an operating system 221, a computer program 222, etc., and the storage method can be temporary storage or permanent storage.

[0100] The operating system 221 is used to manage and control the hardware devices and computer program 222 on the electronic device 20, and can be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program capable of implementing the method for determining aberration parameters of a transmission electron microscope executed by the electronic device 20 as disclosed in any of the aforementioned embodiments, the computer program 222 can further include computer programs capable of implementing other specific tasks.

[0101] Furthermore, this application discloses a computer-readable storage medium for storing a computer program; wherein, when executed by a processor, the computer program implements the aforementioned method for determining aberration parameters of a transmission electron microscope. The specific steps of this method can be found in the corresponding contents disclosed in the aforementioned embodiments and will not be further described here.

[0102] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from the other embodiments. Reference can be made to the descriptions of the identical or similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple, and the relevant parts can be referred to the descriptions of the methods.

[0103] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0104] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0105] Finally, it should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or device comprising the element.

[0106] The above is a detailed introduction to the technical solution provided by the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for those skilled in the art, according to the ideas of the present application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A method for determining aberration parameters of a transmission electron microscope, characterized in that: include: Acquire an original training data set, and acquire an original data set to be measured using a target transmission electron microscope; the original training data set includes first Thorne ring images and first contrast transfer function parameters corresponding to each of the first Thorne ring images; the first contrast transfer function parameters include third-order spherical aberration, defocus, angle of second-order astigmatism, and magnitude of second-order astigmatism; the original data set to be measured includes second Thorne ring images for which contrast transfer function parameters are to be determined; Processing all the first Thorne ring images in the original training data set to obtain binary training images corresponding to the first Thorne ring images, and processing all the second Thorne ring images in the original to-be-measured data set to obtain binary target images corresponding to the second Thorne ring images; A target training set and a target validation set are constructed based on the binarized training images and the corresponding simulated labels, and a target parameter determination model is trained using the target training set and the target validation set; the simulated labels are the first contrast transfer function parameters corresponding to the binarized training images, and the target parameter determination model is a convolutional neural network model for determining contrast transfer function parameters corresponding to Thorne ring images; The binarized target image is input into the target parameter determination model to determine target parameters, so as to optimize the aberration corrector of the target transmission electron microscope based on the target parameters; wherein the target parameters are second contrast transfer function parameters corresponding to the second Thorne ring image.

2. The method for determining aberration parameters of a transmission electron microscope according to claim 1, wherein: The obtaining of the original training data set includes: A preset number of first Thorne ring images are acquired using a preset transmission electron microscope with known contrast transfer function parameters, and an original training data set is constructed using the first Thorne ring images and the first contrast transfer function parameters corresponding to each of the first Thorne ring images.

3. The method for determining aberration parameters of a transmission electron microscope according to claim 1, wherein: The obtaining of the original training data set includes: determining a first contrast transfer function parameter; Simulating an imaging process of a transmission electron microscope using the first contrast transfer function parameters to obtain a first Thorne ring image corresponding to the first contrast transfer function parameters; The original training data set is constructed using the first Thorne ring image and the corresponding first contrast transfer function parameters.

4. The method for determining aberration parameters of a transmission electron microscope according to claim 3, wherein: The simulating the imaging process of a transmission electron microscope using the first contrast transfer function parameters to obtain a first Thorne ring image corresponding to the first contrast transfer function parameters includes: determining an aberration function corresponding to the first contrast transfer function parameter using the aberration parameter of the first transmission electron microscope; determining a one-dimensional contrast transfer function corresponding to the aberration function based on the aberration function; The one-dimensional contrast transfer function is expanded in two dimensions to obtain a first Thorne ring image corresponding to the first contrast transfer function parameters.

5. The method for determining aberration parameters of a transmission electron microscope according to claim 1, wherein: The processing of all the first Thorne ring images in the original training data set to obtain binary training images corresponding to the first Thorne ring images, and processing of all the second Thorne ring images in the original to-be-measured data set to obtain binary target images corresponding to the second Thorne ring images, includes: Performing grayscale processing and noise reduction processing on all the first Thorne ring images in the original training data set to obtain first noise-reduced images; performing thresholding processing on the first denoised image based on a first preset threshold to obtain a binary training image corresponding to the first Thorne ring image; Performing grayscale processing and noise reduction processing on all the second Thorne ring images in the original data set to be measured to obtain second noise-reduced images; The second denoised image is thresholded based on a second preset threshold to obtain a binarized target image corresponding to the second Thorne ring image.

6. The method for determining aberration parameters of a transmission electron microscope according to claim 1, wherein: The step of constructing a target training set and a target verification set based on the binarized training images and the corresponding simulated labels, and using the target training set and the target verification set to train a target parameter determination model includes: Constructing a target training set and a target verification set based on a preset ratio using the binarized training images and corresponding simulated labels; Using the target training set, an initial parameter determination model is trained based on preset training parameters to obtain a trained parameter determination model; the preset training parameters include a learning rate, a batch size, and a number of training rounds; and the initial parameter determination model is a convolutional neural network model; After each iteration is completed, the model parameters of the current post-training parameter determination model are recorded, and the target verification set is input into the current post-training parameter determination model to obtain an error convergence curve; If the current number of iteration rounds is equal to the number of training rounds, target model parameters are determined from all the model parameters based on the error convergence curve, and the trained parameter determination model corresponding to the target model parameters is determined as the target parameter determination model.

7. The method for determining aberration parameters of a transmission electron microscope according to any one of claims 1 to 6, characterized in that: The target parameters determine that the activation function of the last fully connected layer of the model is a linear activation function.

8. A device for determining aberration parameters of a transmission electron microscope, characterized in that: include: a data set acquisition module, configured to acquire an original training data set and to acquire an original data set to be measured using a target transmission electron microscope; the original training data set includes first Thorne ring images and first contrast transfer function parameters corresponding to each of the first Thorne ring images; the first contrast transfer function parameters include third-order spherical aberration, defocus, angle of second-order astigmatism, and magnitude of second-order astigmatism; the original data set to be measured includes second Thorne ring images for which contrast transfer function parameters are to be determined; an image processing module, configured to process all the first Thorne's ring images in the original training data set to obtain binary training images corresponding to the first Thorne's ring images, and to process all the second Thorne's ring images in the original to-be-measured data set to obtain binary target images corresponding to the second Thorne's ring images; a model training module, configured to construct a target training set and a target validation set based on the binarized training images and corresponding simulated labels, and to train a target parameter determination model using the target training set and the target validation set; wherein the simulated labels are the first contrast transfer function parameters corresponding to the binarized training images, and the target parameter determination model is a convolutional neural network model for determining contrast transfer function parameters corresponding to Thorne ring images; A parameter determination module is used to input the binarized target image into the target parameter determination model to determine target parameters, so as to optimize the aberration corrector of the target transmission electron microscope based on the target parameters; wherein the target parameters are second contrast transfer function parameters corresponding to the second Thorne ring image.

9. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the method for determining aberration parameters of a transmission electron microscope according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that Used to store a computer program, wherein when the computer program is executed by a processor, the method for determining the aberration parameters of a transmission electron microscope according to any one of claims 1 to 7 is implemented.

Images