Method, electronic device, and storage medium for field-of-view stitching in a charged particle beam system

Abstract

The present application discloses a method, an electronic device, and a storage medium for field stitching in a charged particle beam system. The method includes: constructing a first relationship model between an optical navigation device and a magnetic deflection device at different working distances based on a first calibration; performing resolution processing on a first image using a preprocessing model to match a second image captured by the magnetic deflection device; forming a first field-of-view switching point from the optical navigation device to the magnetic deflection device according to the first relationship model, the matched first image, and the second image; constructing a second relationship model between the magnetic deflection device and an electric deflection device based on a second calibration; forming a second field-of-view switching point from the magnetic deflection device to the electric deflection device according to the second relationship model; and performing corresponding field compensation based on the first field-of-view switching point and the second field-of-view switching point. By using the solution of the present application, seamless switching of image resolution in different modes can be achieved, and the overall performance of the charged particle beam system can be improved.

Description

Technical Field

[0001] This application generally relates to the field of field stitching technology in charged particle beam systems. More specifically, this application relates to a method, an electronic device, and a storage medium for field stitching in a charged particle beam system. Background Art

[0002] With the continuous improvement of the performance requirements of charged particle beam systems in application scenarios such as scientific research and industrial inspection, how to optimize the imaging function of charged particle beam systems and improve operation convenience has become a key issue. Existing charged particle beam systems mainly rely on magnetic deflection and electric deflection technologies to meet different imaging requirements. Through magnetic deflection, the large-field requirement can be met, enabling operators to observe a larger sample area and obtain overall information; through electric deflection, the high-resolution requirement can be met, clearly presenting the microscopic details of the sample. During the actual operation process, users often need to frequently switch the imaging method in different deflection modes or optical modes to achieve a comprehensive observation from macroscopic to microscopic.

[0003] For example, a system that can seamlessly switch from an optical navigation image to a scanning electron microscope image, realizing a seamless switch of image resolution from hundreds of millimeters to a few nanometers, is of great significance for improving the operation convenience and imaging quality of charged particle beam systems. Such a system allows users to quickly transition from a macroscopic overview to microscopic detail observation without complex mode switching operations during the operation process, greatly improving work efficiency. At the same time, in terms of imaging quality, seamless switching can avoid image information loss or misalignment caused by resolution jumps, ensuring the integrity and accuracy of the image. However, there is currently no effective method in the prior art to achieve seamless switching between optical navigation images and scanning electron microscope images and continuous adjustment of resolution. This results in difficulties in image stitching and resolution mismatch when switching from an optical navigation image to a scanning electron microscope image, making it impossible for users to accurately correspond the sample position determined in the optical navigation with the scanning electron microscope image; when switching between different deflection scanning modes, the central position of the image is prone to shift, and the angle is also difficult to accurately calibrate, seriously affecting the observation and analysis results.

[0004] In view of this, there is an urgent need to provide a solution for field stitching in a charged particle beam system to achieve seamless switching of image resolution in different modes and improve the overall performance of the charged particle beam system. Summary of the Invention

[0005] To solve at least one or more of the above-mentioned technical problems, this application proposes a solution for field stitching in a charged particle beam system in multiple aspects.

[0006] In a first aspect, the present application provides a method for field stitching in a charged particle beam system, where the charged particle beam system includes an optical navigation device, a magnetic deflection device, and an electric deflection device, and the method includes: based on a first calibration, constructing a first relationship model between the optical navigation device and the magnetic deflection device at different working distances; using a preprocessing model to perform resolution processing on a first image captured by the optical navigation device to match a second image captured by the magnetic deflection device; forming a first field-of-view switching point from the optical navigation device to the magnetic deflection device according to the first relationship model, the matched first image, and the second image; based on a second calibration, constructing a second relationship model between the magnetic deflection device and the electric deflection device; forming a second field-of-view switching point from the magnetic deflection device to the electric deflection device according to the second relationship model; performing corresponding field compensation based on the first field-of-view switching point and the second field-of-view switching point to achieve field stitching in the charged particle beam system.

[0007] In one embodiment, the first calibration is performed by the following operations: moving a sample stage, where a first target and a second target are marked on the sample stage, and the first target is located at the center of the sample stage; using the optical navigation device to capture a plurality of first images of the sample stage to calibrate the distance relationship between the first image and the sample stage; based on the distance relationship and using the magnetic deflection device to capture a second image of the sample stage at a target working distance to calibrate the angular relationship between the first image and the second image at the target working distance.

[0008] In another embodiment, using the optical navigation device to capture a plurality of first images of the sample stage to calibrate the distance relationship between the first image and the sample stage includes: obtaining the actual position of the sample stage corresponding to each first image; calculating the moving distances of the first target and the second target on the corresponding first image; determining a distance matrix according to the actual position and the moving distances to calibrate the distance relationship between the first image and the sample stage.

[0009] In another embodiment, based on the distance relationship and using the magnetic deflection device to capture a second image of the sample stage at a target working distance to calibrate the angular relationship between the first image and the second image at the target working distance includes: determining the first positions of the first target and the second target on the first image at the target working distance based on the distance relationship; obtaining the second positions of the first target and the second target on the second image at the target working distance; calibrating the angular relationship between the first image and the second image at the target working distance according to the first positions and the second positions.

[0010] In another embodiment, constructing a first relationship model between the optical navigation device and the magnetic deflection device at different working distances based on the first calibration includes: moving the sample stage and calculating the angular deviation of a second image captured by the magnetic deflection device at different working distances; determining the relative angular difference between the optical navigation device and the magnetic deflection device based on the angular relationship of the first calibration and the angular deviation, so as to construct the first relationship model between the optical navigation device and the magnetic deflection device at different working distances.

[0011] In another embodiment, moving the sample stage and calculating the angular deviation of the second image captured by the magnetic deflection device at different working distances includes: moving the sample stage and calculating the image deviation of the second image captured by the magnetic deflection device at the target feature region before and after changing the working distance; adjusting the deflector of the magnetic deflection device according to the image deviation and a preset threshold until the image deviation is less than the preset threshold, obtaining the corresponding deflection angle, so as to calculate the angular deviation of the second image captured by the magnetic deflection device at different working distances.

[0012] In another embodiment, the preprocessing model includes a downsampling module and an upsampling module, and using the preprocessing model to perform resolution processing on the first image captured by the optical navigation device includes: performing a downsampling operation on the first image captured by the optical navigation device using the downsampling module to obtain downsampled features; performing an upsampling operation on the downsampled features using the upsampling module and fusing the corresponding downsampled features during the upsampling operation to obtain upsampled features; and superimposing the final upsampled features on the first image to obtain a target image, so as to perform resolution processing on the first image captured by the optical navigation device.

[0013] In another embodiment, the preprocessing model is a Restormer model.

[0014] In another embodiment, constructing a second relationship model between the magnetic deflection device and the electric deflection device based on the second calibration includes: calculating the magnification and angular difference between the deflector of the magnetic deflection device and the deflector of the electric deflection device based on the second calibration, so as to construct the second relationship model between the magnetic deflection device and the electric deflection device.

[0015] In another embodiment, calculating the magnification and the angular difference between the deflector of the magnetic deflection device and the deflector of the electric deflection device based on the second calibration includes: respectively identifying the first grid number and the second grid number corresponding to the calibration sample under the magnetic deflection device and the electric deflection device; adjusting the magnification gain factor of the deflector of the magnetic deflection device so that the first grid number is consistent with the second grid number to calculate the magnification between the deflector of the magnetic deflection device and the deflector of the electric deflection device; identifying the angular deviation between the corresponding grids of the magnetic deflection device and the electric deflection device at the magnification to calculate the angular difference between the deflector of the magnetic deflection device and the deflector of the electric deflection device.

[0016] In another embodiment, it further includes: setting a third field-of-view switching point and a fourth field-of-view switching point; performing corresponding field-of-view compensation based on the first field-of-view switching point, the second field-of-view switching point, the third field-of-view switching point, and the fourth field-of-view switching point to achieve field-of-view stitching in the charged particle beam system.

[0017] In another embodiment, the third field-of-view switching point and the fourth field-of-view switching point are the reverse field-of-view switches of the first field-of-view switching point and the second field-of-view switching point.

[0018] In a second aspect, the present application provides an electronic device, including: a processor; and a memory, where program instructions for field-of-view stitching in a charged particle beam system are stored, and when the program instructions are executed by the processor, the electronic device implements one or more embodiments in the foregoing first aspect.

[0019] In a third aspect, the present application provides a computer-readable storage medium, on which computer-readable instructions for field-of-view stitching in a charged particle beam system are stored, and when the computer-readable instructions are executed by one or more processors, one or more embodiments in the foregoing first aspect are implemented.

[0020] Through the solution for field stitching in a charged particle beam system provided above, embodiments of the present application calibrate to construct a first relationship model between an optical navigation device and a magnetic deflection device, and a second relationship model between the magnetic deflection device and an electric deflection device, and perform resolution processing on a first image captured by the optical navigation device to match a second image captured by the magnetic deflection device, filling the gap from the optical navigation device to magnetic and electric deflection scanning. By the foregoing calibration, a first field switching point and a second field switching point are formed. By performing corresponding field compensation at the corresponding field switching points, the stitching angle between the navigation camera and magnetic and electric deflection scanning is accurate, ensuring the image stitching accuracy, avoiding angle deviation or misalignment after image stitching, providing an accurate image basis for subsequent analysis, realizing seamless switching of image resolution in different modes, and improving the overall performance of the charged particle beam system. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features, and advantages of the exemplary embodiments of the present application will become readily understood. In the drawings, several embodiments of the present application are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0022] Figure 1 is an exemplary schematic diagram showing image switching in a charged particle beam system;

[0023] Figure 2 is an exemplary flowchart showing a method for field stitching in a charged particle beam system according to an embodiment of the present application;

[0024] Figure 3 is an exemplary schematic diagram showing a sample stage for performing a first calibration according to an embodiment of the present application;

[0025] Figure 4 is an exemplary schematic diagram showing using an optical navigation device to capture a sample stage according to an embodiment of the present application;

[0026] Figure 5 is an exemplary schematic diagram showing respective recognition targets on a first image and a second image according to an embodiment of the present application;

[0027] Figure 6 is an exemplary flowchart showing establishing an angular relationship at different working distances according to an embodiment of the present application;

[0028] Figure 7 is an exemplary schematic diagram showing the angular relationship at different working distances according to an embodiment of the present application;

[0029] Figure 8It is an exemplary schematic diagram showing the resolution processing of the first image captured by the optical navigation device according to an embodiment of the present application;

[0030] Figure 9 It is an exemplary flowchart showing the calculation of the amplification gain multiple and the angle difference between two deflectors according to an embodiment of the present application;

[0031] Figure 10 It is an exemplary schematic diagram of identifying a sample grid according to an embodiment of the present application;

[0032] Figure 11 It is an exemplary schematic diagram of image switching for increasing the field of view switching point according to an embodiment of the present application;

[0033] Figure 12 It is an exemplary structural block diagram of an electronic device according to an embodiment of the present application. Detailed implementation manners

[0034] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts fall within the protection scope of the present application.

[0035] It should be understood that the terms "including" and "comprising" used in the specification and claims of the present application indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0036] It should also be understood that the terms used in the specification of the present application are only for the purpose of describing specific embodiments, and are not intended to limit the present application. As used in the specification and claims of the present application, unless the context clearly indicates otherwise, the singular forms "a", "an", and "the" are intended to include the plural forms. It should be further understood that the term " / and / " used in the specification and claims of the present application refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0037] As used in this specification and the claims, the term "if" can be interpreted as "when" or "once" or "in response to determining" or "in response to detecting" depending on the context. Similarly, the phrase "if determined" or "if [described condition or event] is detected" can be interpreted as meaning "once determined" or "in response to determining" or "once [described condition or event] is detected" or "in response to detecting [described condition or event]" depending on the context.

[0038] Figure 1 is an exemplary schematic diagram showing image switching in a charged particle beam system. As Figure 1 shown, the charged particle beam system may include an optical navigation device (corresponding to OM), a magnetic deflection device (corresponding to magnetic deflection), and an electric deflection device (corresponding to electric deflection), where the magnetic deflection device and the electric deflection device together form a magneto-electric composite scanning deflection focusing system, such as a scanning electron microscope. In an actual application scenario, the entire switching process can be divided into two segments. The first segment is from the optical navigation device to the magnetic deflection device, i.e., OM → magnetic. The second segment is from the magnetic deflection device to the electric deflection device, i.e., magnetic → electric. According to the description of the background technology, when switching from an optical navigation image to a scanning electron microscope image, there are often problems such as difficult image stitching and mismatched resolutions, making it impossible for the user to accurately correspond the sample position determined in the optical navigation with the scanning electron microscope image. Additionally, when switching from the magnetic deflection device to the electric deflection device, the center position of the image is prone to shift, and the angle is difficult to accurately calibrate, seriously affecting the observation and analysis results. There is a lack of an effective method in the prior art to achieve seamless switching between the optical navigation image and the scanning electron microscope image and continuous adjustment of the resolution.

[0039] As previously mentioned, the charged particle beam system may include an optical navigation device, a magnetic deflection device, and an electric deflection device. In some embodiments, the optical navigation device may include a navigation lens and an industrial camera. In the implementation scenario, in order to ensure that the optical navigation image can be stitched with the magnetic and electric deflection images, it is first necessary to select the optical navigation device to adapt to the magnetic deflection device and the electric deflection device. For example, the theoretical resolution of the optical navigation device needs to reach 30um, so that after the image of the optical navigation device is cropped and processed by a super-resolution algorithm, its resolution can match the images of the magnetic deflection device and the electric deflection device under the application conditions.

[0040] In addition, the field of view ("FOV") of the optical navigation device needs to meet the requirements of the user for viewing and finding samples. Specifically, the FOV should at least cover the entire area where the sample tray is located. It can be understood that, with the size and pixels of the camera's photosensitive element fixed, a larger FOV means larger pixel points and lower resolution. Therefore, the FOV needs to be appropriate and not too large. In addition, try to ensure that the image plane resolution of the navigation lens is the same as that of the industrial camera chip (CMOS / CCD), that is, the pixel size is the same.

[0041] Based on the above requirements, relevant parameters can be calculated based on the formulas related to geometric optics: 、 and to select the appropriate optical navigation device. Among them, represents the image distance, l represents the object distance, represents the lens focal length, θ represents the field of view angle, β represents the lateral magnification, represents the image height, and y represents the object height. That is, after clarifying the object height y, object distance l, and object resolution required by the navigation camera, the required field of view range can be deduced, or the object resolution can be calculated based on the pixel size to see if it meets the requirements. For example, for a 1 / 1.7" CMOS chip with a target surface size of 5.7 mm and an object height of 90 mm, the lateral magnification can be calculated as 1 / 15.8 according to the above formula; the pixel y' is 1.85 μm, and the object resolution is calculated as 29.2 μm (1.85 × 15.8), which meets the requirement of the object space resolution of 30 μm. Therefore, after clarifying the requirements for the object height, object distance, and object resolution parameters, the field of view angle, pixel size, focal length, etc. calculated according to the above-mentioned geometric optics related formulas are used to select the navigation camera accordingly.

[0042] In some other embodiments, the magnetic deflection device and the electric deflection device each include corresponding deflectors. Among them, the electric deflection device is realized by an electrostatic deflector. The electrostatic deflector includes a number of non-connected metal electrodes distributed on the same cylindrical surface. Its basic principle is to control the deflection of the electron beam by applying a voltage signal to the pole piece. To achieve this effect, the size arrangement and excitation voltage distribution of the deflector electrodes conform to a specific law to generate an approximately uniform bipolar deflection electric field with a specific intensity inside the cylindrical surface.

[0043] The magnetic deflection device is realized by an electromagnetic deflector, which refers to a number of electromagnetic coils distributed axially symmetrically. Its basic principle is to control the deflection of the electron beam by applying a current signal to the coil. To achieve this effect, the coil shape may be rectangular, saddle-shaped or annular, and its structure may be hollow or with an iron core, aiming to generate an approximately uniform secondary deflection electric field with a specific intensity inside the electromagnetic deflector.

[0044] In some other embodiments, the charged particle beam system may further include, for example, an electron lens, which refers to an electrostatic or electromagnetic lens with a circularly symmetric structure and used to converge an electron beam. In an implementation scenario, several groups of electrostatic deflectors, several groups of electromagnetic deflectors, and several groups of electron lenses are coaxially distributed along the main axis of the electron beam or ion beam at specific positions. When the user changes the magnification (i.e., enlarges the field of view), the deflectors and lenses will automatically switch the combination mode, that is, control the on / off of the scanning excitation signal on each group of deflectors and the on / off of the excitation signal on the electron lens.

[0045] For example, in the case of the lowest magnification (i.e., the largest field of view), the charged particle beam system of the embodiment of the present application will adopt a combination of lenses at a higher position and magnetic deflectors to form a post-lens deflection system; when the magnification is the highest (i.e., the smallest field of view), the charged particle beam system of the embodiment of the present application will adopt a combination of lenses at a lower position and electrostatic deflectors to form a double pre-lens deflection system or an in-lens deflection system. At this time, the overall aberration of the system is the smallest, the resolution is the highest, and the scanning speed is also the fastest. Between the above two cases, there are also several other combinations of deflectors and lenses, and each combination corresponds to the best imaging conditions within a certain field of view range, and these field of view ranges overlap with each other.

[0046] Due to the limitations of the mechanical structure and imaging conditions, the images under different combinations of deflectors and objective lenses may have differences in angle or central position. At the same time, there may also be differences in angle and central position between the optical navigation map and the electron microscope image. By selecting appropriate field of view stitching points, switching different usage conditions, and using image recognition and processing algorithms, and changing parameters such as the scanning direction and the position of the sample stage, the images under different conditions can be corrected and aligned, and thus a seamless switch from a large field of view to high-speed and high-resolution can be achieved. Similarly, a seamless switch from the optical navigation map to the electron microscope image can also be achieved.

[0047] Based on this, the present application provides a method for field of view stitching in a charged particle beam system, which can achieve seamless switching of image resolution in different modes and improve the overall performance of the charged particle beam system.

[0048] The following will describe in detail the specific implementation manners of the present application with reference to the accompanying drawings.

[0049] Figure 2 is an exemplary flowchart showing a method 200 for field of view stitching in a charged particle beam system according to an embodiment of the present application. As Figure 2As shown in [figure], method 200 includes step S201: based on the first calibration, construct a first relationship model between the optical navigation device and the magnetic deflection device at different working distances; step S202: use the preprocessing model to perform resolution processing on the first image captured by the optical navigation device to match the second image captured by the magnetic deflection device; step S203: form a first field-of-view switching point from the optical navigation device to the magnetic deflection device according to the first relationship model, the matched first image and the second image; step S204: based on the second calibration, construct a second relationship model between the magnetic deflection device and the electric deflection device; step S205: form a second field-of-view switching point from the magnetic deflection device to the electric deflection device according to the second relationship model; step S206: perform corresponding field-of-view compensation based on the first field-of-view switching point and the second field-of-view switching point to achieve field-of-view stitching in the charged particle beam system.

[0050] First, at step S201, based on the first calibration, construct a first relationship model between the optical navigation device and the magnetic deflection device at different working distances. It can be understood that at different working distances, due to different objective lens excitations, there will be an angular difference between the first image of the optical navigation device and the second image captured by the magnetic deflection device; or the second image captured by the magnetic deflection device has a rotation angle. Thus, by establishing the first relationship model between the optical navigation device and the magnetic deflection device through the first calibration, the existing angular difference can be determined, so as to compensate for the angular difference by adjusting the working distance during field-of-view stitching to achieve field-of-view stitching.

[0051] In some embodiments, the first calibration can be performed through the following operations: move the sample stage, where the first target and the second target are marked on the sample stage, and the first target is located at the center of the sample stage. Then, use the optical navigation device to capture multiple first images of the sample stage to calibrate the distance relationship between the first image and the sample stage, and further, based on the distance relationship and use the magnetic deflection device to capture the second image of the sample stage at the target working distance to calibrate the angular relationship between the first image and the second image at the target working distance. That is, first determine the distance relationship between the first image and the sample stage by capturing with the optical navigation device, and then calibrate the angular relationship between the first image and the second image at the target working distance by combining the capture with the magnetic deflection device and the distance relationship.

[0052] In some implementation scenarios, the aforementioned first target and second target can include, but are not limited to, a sample in the shape of a circle or other features, and the first target is located at the center of the sample stage, and the second target can be located near the first target. When moving the sample stage and using the optical navigation device to capture multiple first images of the sample stage, the corresponding first image can be captured when the sample stage is in the original position (not moved), the corresponding first image can be captured when the sample stage is moved along the x-axis direction, and the corresponding first image can be captured when the sample stage is moved along the y-axis direction.

[0053] In some embodiments, the actual position of the sample stage corresponding to each first image can be obtained, the moving distances of the first target and the second target on the corresponding first image can be calculated, and then the distance matrix can be determined based on the actual position and the moving distances to calibrate the distance relationship between the first image and the sample stage. As an example, assume that the matrix composed of the actual positions of the sample stage corresponding to three first images is , and the matrix composed of the moving distances of the first and second targets on the first image is . In this case, the distance matrix can be calculated based on to calibrate the distance relationship between the first image and the sample stage.

[0054] In some embodiments, the first positions of the first target and the second target on the first image at the target working distance can be determined based on the distance relationship, then the second positions of the first target and the second target on the second image at the target working distance can be obtained, and the angular relationship between the first image and the second image at the target working distance can be calibrated according to the first positions and the second positions. In the implementation scenario, the working distance corresponds to the distance between the electron beam source and the sample stage, and thus the actual position S of the sample stage at the target working distance can be determined. By combining the above distance relationship , the first positions of the first target and the second target on the first image at the target working distance can be obtained. For example, the first positions of the first target and the second target on the first image at the target working distance are respectively denoted as (x’1, y’1), (x’2, y’2).

[0055] Furthermore, the second positions of the first target and the second target on the second image at the target working distance are obtained, for example, denoted as (x1, y1), (x2, y2) respectively. In this scenario, the angular relationship between the first image and the second image at the target working distance can be calibrated by polar coordinates , where . Thus, the aforementioned first calibration is completed.

[0056] In some embodiments, the sample stage is moved, the angular deviation of the second image captured by the magnetic deflection device at different working distances is calculated, and the relative angular difference between the optical navigation device and the magnetic deflection device is determined based on the angular relationship of the first calibration and the angular deviation to construct the first relationship model between the optical navigation device and the magnetic deflection device at different working distances. More specifically, in some embodiments, the sample stage is moved, the image deviation of the second image captured by the magnetic deflection device at the target feature region before and after changing the working distance is calculated, the deflector of the magnetic deflection device is adjusted according to the image deviation and the preset threshold until the image deviation is less than the preset threshold, and the corresponding deflection angle is obtained to calculate the angular deviation of the second image captured by the magnetic deflection device at different working distances.

[0057] It can be understood that during actual user operation, the sample stage needs to be moved to different positions along the Z-axis (electron beam direction), and the corresponding working distances are different. Based on this, the angular relationships at different working distances are established. As an example, calibrate the angles at multiple working distances. For example, calibrate the angles every 5 mm between 5 mm and 50 mm, and select a target feature area (such as a solder ball sample). By moving the sample stage, match the images before and after the movement, and continuously iterate until the deviation between the images before and after the movement is less than a preset threshold, and then calculate the current deflection angle. Then, change to the next calibration point and calculate the current deflection angle in the same way. Finally, fit the calibration points and angles into a curve relationship formula, that is, the angular deviation. Combining the aforementioned angular relationship based on the first calibration, the relative angular difference can be obtained. More details about constructing the first relationship model between the optical navigation device and the magnetic deflection device at different working distances based on the first calibration will be described later in conjunction with Figures 3 - 7 detailed description.

[0058] Next, at step S202, use the preprocessing model to perform resolution processing on the first image captured by the optical navigation device to match the second image captured by the magnetic deflection device. Based on this, through super-resolution, the pixel size of the first image captured by the optical navigation device can be stitched with the field of view of the second image captured by the magnetic deflection device.

[0059] In some embodiments, the preprocessing model may include a downsampling module and an upsampling module. First, perform a downsampling operation on the first image captured by the optical navigation device using the downsampling module to obtain downsampled features. Then, use the upsampling module to perform an upsampling operation on the downsampled features, and fuse the corresponding downsampled features during the upsampling operation to obtain upsampled features. And superimpose the final upsampled features on the first image to obtain the target image, so as to perform resolution processing on the first image captured by the optical navigation device. In some embodiments, the preprocessing model may be, for example, the Restormer model.

[0060] In some implementation scenarios, the downsampling module may include a convolutional layer and a transformer layer, and the upsampling module may include a transformer layer, a feature fusion layer, and a transposed convolutional layer. In some other implementation scenarios, the upsampling module may further include a feed-forward network (such as GDFN) and an attention layer (such as MDTA). In some other implementation scenarios, before inputting the first image into the preprocessing model, preprocessing operations such as image cropping and image size adjustment may also be performed on the first image. More details about the resolution processing will be described later in conjunction with Figure 8 detailed description.

[0061] Further, at step S203, a first field-of-view switching point from the optical navigation device to the magnetic deflection device is formed according to the first relationship model, the matched first image, and the second image. That is, according to the first relationship model, the matched first image, and the second image, the first field-of-view switching of the optical navigation device to the magnetic deflection device can be realized.

[0062] At step S204, based on the second calibration, a second relationship model between the magnetic deflection device and the electric deflection device is constructed. In some embodiments, the magnification and the angular difference between the deflector of the magnetic deflection device and the deflector of the electric deflection device are calculated to construct the second relationship model between the magnetic deflection device and the electric deflection device.

[0063] More specifically, in some embodiments, the first grid number and the second grid number corresponding to the calibration sample under the magnetic deflection device and the electric deflection device are respectively identified, and the magnification gain of the deflector of the magnetic deflection device is adjusted so that the first grid number is consistent with the second grid number to calculate the magnification between the deflector of the magnetic deflection device and the deflector of the electric deflection device. Further, the angular deviation between the grids corresponding to the magnetic deflection device and the electric deflection device at the magnification is identified to calculate the angular difference between the deflector of the magnetic deflection device and the deflector of the electric deflection device.

[0064] It can be understood that there is a linear relationship between the field-of-view size and the magnification gain of the deflector. Therefore, according to the multiple values of the actual field-of-view sizes of the magnetic deflection device and the electric deflection device, the magnification gain of the deflector can be adjusted by the corresponding multiple to keep the fields-of-view of the two deflectors consistent. At the same time, the grid deflection angle is identified to calculate the angular deviation between the two deflectors, so as to construct the second relationship model between the magnetic deflection device and the electric deflection device. More details about constructing the second relationship model will be described later in combination with Figure 9 and Figure 10 will be described in detail.

[0065] Further, at step S205, a second field-of-view switching point from the magnetic deflection device to the electric deflection device is formed according to the second relationship model. That is, according to the second relationship model, the second field-of-view switching of the magnetic deflection device to the electric deflection device can be realized. Finally, at step S206, corresponding field-of-view compensation is performed based on the first field-of-view switching point and the second field-of-view switching point to achieve field-of-view stitching in the charged particle beam system.

[0066] Specifically, according to the relative angular difference, magnification, and angular deviation determined by the first relationship model and the second relationship model at the first field-of-view switching point and the second field-of-view switching point respectively, when in use, the deflector will be automatically switched at the switching point by increasing or decreasing the magnification, realizing seamless magnification from a millimeter-level field-of-view to a nanometer-level field-of-view. Among them, at the first field-of-view switching point, resolution processing is also performed on the first image to be in the super-resolution region (such as the above Figure 1As shown in [Figure], the missing part is completed to make the image of the optical navigation device coincide with the image scanned by the magnetic deflection device, improving the overall performance of the charged particle beam system.

[0067] Figure 3 is an exemplary schematic diagram showing a sample stage for performing the first calibration according to an embodiment of the present application. As Figure 3 shown on the left in [Figure], it is a top view of the sample stage, and the right side shown is a side view of the sample stage. The sample stage is a standard nail stage, and the first target 1 and the second target 2 with holes having a diameter of 1 mm are exemplarily shown on the sample stage, where the first target 1 is located at the center of the sample stage, and the second target 2 is 2 mm away from the first target 1. In the implementation scenario, first, the optical navigation device is used to photograph the sample stage to calibrate the distance relationship between the first image of the optical navigation device and the sample stage.

[0068] Figure 4 is an exemplary schematic diagram showing the use of an optical navigation device to photograph a sample stage according to an embodiment of the present application. As Figure 4 shown at the lower left in [Figure], the sample stage is in the original position. The three first images taken by the optical navigation device when the sample stage is moved 10 mm along the x-axis direction as shown on the lower right in the figure and when the sample stage is moved 10 mm along the y-axis direction as shown at the upper part in the figure are shown. It can be understood that the x-axis and the y-axis are two directions perpendicular to the electron beam direction (i.e., the z-axis) and in the same plane. In the implementation scenario, based on the three first images taken, the actual position S of the corresponding sample stage is obtained. Then, according to the first images at different positions, the pixel distance by which the first target 1 and the second target 2 move in the image is calculated. For example, after the sample stage is moved 10 mm along the x-axis direction from the original position, the difference between the pixel coordinates of the first target 1 and the second target 2 on the first image at the original position and the pixel coordinates of the first target 1 and the second target 2 on the first image after moving 10 mm along the x-axis direction is the moving distance.

[0069] Similarly, the difference between the pixel coordinates of the first target 1 and the second target 2 on the first image at the original position and the pixel coordinates of the first target 1 and the second target 2 on the first image after moving 10 mm along the y-axis direction is calculated; or the difference between the pixel coordinates of the first target 1 and the second target 2 on the first image after moving 10 mm along the x-axis direction and the pixel coordinates of the first target 1 and the second target 2 on the first image after moving 10 mm along the y-axis direction is calculated, and a set of moving distances D can be obtained. According to the calculated distance matrix , the distance relationship between the first image and the sample stage is calibrated.

[0070] Based on the above distance relationship , it is possible to map the points on the first image captured by the optical navigation device to the actual sample stage position; or based on the sample stage position, map to the points on the first image captured by the optical navigation device. Based on this, the sample stage position can be determined at the target working distance, so that the position of the target on the first image captured by the optical navigation device can be identified. In addition, according to the pixel coordinates on the second image captured by the magnetic deflection device, the position of the target on the second image can also be identified, for example Figure 5 as shown

[0071] Figure 5 FIG. is an exemplary schematic diagram showing the identification of the target on the first image and the second image according to an embodiment of the present application. As shown Figure 5 on the left side in, the first target 1 and the second target 2 identified on the first image can automatically identify their positions according to the above distance relationship , for example, denoted as (x'1, y'1) and (x'2, y'2) respectively Figure 5 In the middle, the first target 1 and the second target 2 identified on the second image can obtain their corresponding positions according to the pixel coordinates, for example, denoted as (x1, y1) and (x2, y2) respectively. In this scenario, based on the above angular relationship the angle between the first image and the second image at the target working distance can be calibrated , for example Figure 5 as shown on the right side in

[0072] According to the foregoing, there will also be a rotation angle for the second image at different working distances, thereby establishing an angular relationship at different working distances. Subsequently, by combining the calibrated angular relationship between the first image and the second image, the angular phase difference between the optical navigation device and the magnetic deflection device can be obtained to construct a first relationship model

[0073] Figure 6 FIG. is an exemplary flowchart showing the establishment of the angular relationship at different working distances according to an embodiment of the present application. As shown Figure 6 in, at step S601, the sample stage is moved to the calibration point. For example, every 5 mm between the angles of 5 mm - 50 mm is a calibration point, so that the working distance of the sample stage can reach the 5 mm position. Then, at step S602, a target feature area is selected in the second image, and at step S603, the sample stage Y-axis is moved. As an example, the aforementioned moving distance can be 1 / 5 of the current field of view. Further, at step S604, the image deviation of the second image captured by the magnetic deflection device at the target feature area before and after changing the working distance is calculated. In some implementation scenarios, the image deviation in the current y direction can be calculated by, for example, an image matching algorithm

[0074] At step S605, it is determined whether the image deviation is less than a preset threshold. In some implementation scenarios, the preset threshold may be, for example, 0.1 pixel. When the image deviation is not less than the preset threshold, at step S606, the deflector of the magnetic deflection device is adjusted, and the process returns to step S603 to move the Y-axis again to compare the new image deviation with the preset threshold to adjust the deflector. The specific adjustment method is as follows: if the image deviation is greater than 10 pixels, the deflector angle of the magnetic deflection device is adjusted by 1.5°; if the image deviation is between 3 and 10 pixels, the deflector angle of the magnetic deflection device is adjusted by 0.5°; if the image deviation is less than 3 pixels, the deflector angle of the magnetic deflection device is adjusted by 0.1°. Until the image deviation is less than the preset threshold, at step S607, the current deflection angle is recorded. Then, by changing to the next calibration point, the foregoing steps are repeated to obtain the deflection angles at different working distances.

[0075] Figure 7 FIG. is an exemplary diagram showing the angular relationship at different working distances according to an embodiment of the present application. As Figure 7 shown, the abscissa is the working distance (unit: mm), and the ordinate is the deflection angle (unit: °). It can be seen from the figure that the angular relationship at different working distances is finally fitted to a curve relationship. For example, the dotted line in the figure shows the deflection angles corresponding to different working distances, and the solid line represents the finally fitted curve relationship y = -0.0004x^3 + 0.0537x^2 - 2.467x + 5.6253. In an implementation scenario, in combination with the angular relationship between the first image and the second image calibrated above, the angular phase difference between the optical navigation device and the magnetic deflection device can be obtained. As an example, assume that according to the angular relationship at different working distances above, when the working distance WD = 5 is changed to the working distance WD = 10, the deflection angle corresponding to the second image is 5°; when the working distance WD = 10 is changed to the working distance WD = 15, the deflection angle corresponding to the second image is 3°. Based on the angle of 30° between the first image and the second image at the target working distance WD = 15 obtained from the above first calibration, the angular phase difference between the first image and the second image at the working distance WD = 10 can be calculated to be 27°. In this scenario, by rotating the optical navigation device to compensate for this angular phase difference at the first field-of-view switching point, the first image and the second image can be stitched in the field of view.

[0076] Figure 8 FIG. is an exemplary diagram showing the resolution processing of the first image captured by the optical navigation device according to an embodiment of the present application. As Figure 8As shown in [figure], the preprocessing model may include a downsampling module 801 and an upsampling module 802. The downsampling module 801 includes a convolutional layer 801-1 and multiple Transformer layers 801-2. The upsampling module 802 includes multiple Transformer layers 802-1, a feature fusion layer 802-2, and a transposed convolutional layer 802-3. In addition, the upsampling module 802 may further include a GDFN layer 802-4 and an MDTA layer 802-5. The GDFN layer 802-4 introduces a gating mechanism and depthwise separable convolutions to enhance the feature representation ability and reduce the computational load. The MDTA layer 802-5 introduces depth convolutions and transposed attention to efficiently capture long-range dependencies for image restoration such as image deblurring.

[0077] In an implementation scenario, first, the first image can be cropped into image patches 803 and then enlarged in size to obtain an enlarged image 804. As an example, the first image captured by the optical navigation device is cut into multiple 256×256 image patches and resized to 1024×1024. Then, the enlarged image 804 is input into the preprocessing model. First, it undergoes a downsampling operation through a convolutional layer 801-1 and multiple Transformer layers 801-2 in the downsampling module 801 to obtain downsampled features. Further, it undergoes an upsampling operation through multiple Transformer layers 802-1, a feature fusion layer 802-2, an MDTA layer 802-5, a GDFN layer 802-4, and a transposed convolutional layer 802-3 in the upsampling module 802. Among them, the corresponding-layer downsampled features are fused through the feature fusion layer 802-2 to obtain upsampled features 805. Further, the upsampled features 805 and the enlarged image 804 are superimposed to obtain a target image 806.

[0078] It can be understood that, according to the selection of the optical navigation device, assuming that the field of view size is 122 mm × 91 mm and the size of the first captured image is 4072×3046, the actual pixel size can be calculated as approximately 30 um based on the shortest side 91 mm / 3046×1000≈30 um. When the acceleration voltage of the magnetic deflection device is 30 kv and the working distance is 5 mm, the limit condition is reached. Under this condition, the magnetic scan field of view is the smallest. After actual testing, the minimum field of view is 4.5 mm × 4.5 mm, which can meet the most extreme conditions and cover other acceleration voltages and working distances. Since the field of view stitching of the optical navigation device and the magnetic deflection device needs to ensure the same resolution, but the actual situation does not meet this requirement, the above-mentioned resolution processing operation is implemented in the embodiments of the present application, so that the pixel size of the optical navigation device coincides with the pixel size of the magnetic deflection device. That is, by using the preprocessing model to restore the image details of the first image and reduce the pixel size of the first image. In the case where the pixel size of the first captured image is 30 um, by enlarging the image and keeping the field of view fixed, the pixel size will decrease according to the magnification factor until it coincides with the field of view of the second image captured by the magnetic deflection device. For example, the first image with a size of 4072×3046 captured by the optical navigation device is super-resolved and reconstructed into an image with a size of 28504×21322 by the preprocessing model, and the field of view remains fixed. The current pixel size is calculated to be 4.27 um. This value is smaller than the field of view size of the second image captured by the magnetic deflection device, meeting the condition of overlapping fields of view with the second image.

[0079] Figure 9 It is an exemplary flowchart showing the calculation of the magnification gain factor and the angle difference between two deflectors according to an embodiment of the present application. As Figure 9 shown, at step S901, a standard sample is selected. The standard sample needs to have a known true size and have straight lines or other identical features. Based on this standard sample, at step S902, the charged particle beam system is switched to the magnetic deflection device, and at step S903, the sample grid is recognized by an image recognition algorithm to calculate the field of view size under magnetic deflection. It can be understood that the number of grids corresponds to the field of view size.

[0080] Next, at step S904, the charged particle beam system is switched to the electric deflection device, and at step S905, the sample grid is identified by an image recognition algorithm to calculate the field of view size under electric deflection. Based on the field of view size under magnetic deflection and the field of view size under electric deflection, at step S906, the magnification gain factor is calculated. Specifically, by adjusting the magnification gain factor of the deflector of the magnetic deflection device, the field of view size under magnetic deflection and the field of view size under electric deflection are made consistent. That is, when the corresponding number of grids is the same, the magnification gain factor is obtained. Further, at step S907, the angular deviation is calculated based on the angle between the grid corresponding to the adjusted magnification gain factor and the grid corresponding to the electric deflection, and the second relationship model between the magnetic deflection device and the electric deflection device is obtained.

[0081] Figure 10 is an exemplary schematic diagram of identifying a sample grid according to an embodiment of the present application. As Figure 10 shown on the left side, the original sample grid is shown, and in the middle, the sample grid corresponding to the adjusted magnification gain factor is shown. As can be seen from the figure, the sample grid corresponding to the adjusted magnification gain factor has a rotation. In an implementation scenario, the grid can be grouped by rows and columns, and the grid cross is identified, for example Figure 10 shown on the right side, to calculate the angular deviation.

[0082] Combined with the above description, through the above first and second calibrations, two field of view switching points from the optical navigation device to the electric deflection device can be determined (for example Figure 1 the field of view switching points O1, O2 shown), including the relative angular difference between the optical navigation device and the magnetic deflection device, the angular deviation between the magnetic deflection device and the electric deflection device, and the magnification gain factor. In an implementation scenario, based on the foregoing two field of view switching points, the missing part is complemented by using an image super-resolution reconstruction technique on the first image captured by the optical navigation device by using a preprocessing model, so that the first image coincides with the second image captured by the magnetic deflection device. Further, by adjusting the magnification factor in the system, the system seamlessly switches the magnification between 1 and 1 million times, automatically selects the field of view mode to match the current field of view size, and adjusts parameters such as the relative angular difference and the angular deviation to achieve a visually seamless switch. Based on this, the efficiency of using the charged particle beam system is greatly improved, and by selecting the appropriate field of view size according to the magnification, the image is always ensured to be in the best display effect.

[0083] In some embodiments, the embodiments of the present application can also set a third field of view switching point and a fourth field of view switching point, and perform corresponding field of view compensation based on the first field of view switching point, the second field of view switching point, the third field of view switching point, and the fourth field of view switching point to achieve field of view stitching in the charged particle beam system. Among them, the third field of view switching point and the fourth field of view switching point are the reverse field of view switching of the first field of view switching point and the second field of view switching point.

[0084] Figure 11 is an exemplary schematic diagram of image switching for increasing the field of view switching point according to an embodiment of the present application. As Figure 11 shown, in addition to the first field of view switching point O1 and the second field of view switching point O2, a third field of view switching point O3 and a fourth field of view switching point O4 are also set. That is, the embodiment of the present application adds a deflector redundancy design. By setting four field of view switching points and dividing them into two groups. The first group includes two reference points for the field of view switching from the optical navigation device with an increasing positive magnification to the magnetic deflector device and the field of view switching from the magnetic deflector device to the electric deflector device. The second group includes two reference points for the field of view switching from the electric deflector device to the magnetic deflector device with a decreasing reverse magnification and the field of view switching from the magnetic deflector device to the optical navigation device. In some implementation scenarios, 1.2 times the parameters (amplification gain factor) calibrated in the previous text can be selected as the field of view switching point for the first group of positive adjustments, and 1 to 1.2 times the parameters adjacent to the field of view switching point of the first group can be used as the field of view switching point for the second group of reverse adjustments. Based on this, it is not only possible to prevent the user from repeatedly switching the view mode at the critical point, but also to avoid the problem that due to the magnification being calculated based on the current pixel size and there being an error, it causes repeated automatic switching at a certain magnification during use, affecting the use.

[0085] Figure 12 is an exemplary structural block diagram showing an electronic device 1200 according to an embodiment of the present application. As Figure 12 shown, the electronic device 1200 of the present application may include a processor 1201 and a memory 1202, where the processor 1201 and the memory 1202 communicate through a bus. The memory 1202 stores program instructions for field of view stitching in a charged particle beam system. When the program instructions are executed by the processor 1201, the method steps described in the previous text in combination with the drawings are implemented: based on the first calibration, constructing a first relationship model between the optical navigation device and the magnetic deflector device at different working distances; using a preprocessing model to perform resolution processing on the first image captured by the optical navigation device to match the second image captured by the magnetic deflector device; forming a first field of view switching point from the optical navigation device to the magnetic deflector device according to the first relationship model, the matched first image and the second image; based on the second calibration, constructing a second relationship model between the magnetic deflector device and the electric deflector device; forming a second field of view switching point from the magnetic deflector device to the electric deflector device according to the second relationship model; performing corresponding field of view compensation based on the first field of view switching point and the second field of view switching point to achieve field of view stitching in the charged particle beam system.

[0086] Based on the above description with reference to the accompanying drawings, those skilled in the art can also understand that the embodiments of the present application can also be implemented by software programs. Accordingly, the present application also provides a computer-readable storage medium. The computer-readable storage medium stores computer-readable instructions for field-of-view stitching in a charged particle beam system. When the computer-readable instructions are executed by one or more processors, the method for field-of-view stitching described in the present application in combination with the attached Figure 2 drawings is implemented.

[0087] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0088] It should be noted that although the operations of the method of the present application are described in a specific order in the drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. On the contrary, the order of execution of the steps depicted in the flowchart can be changed. Additionally or alternatively, some steps can be omitted, multiple steps can be combined into one step for execution, and / or one step can be decomposed into multiple steps for execution.

[0089] It should be understood that when terms such as "first", "second", "third", and "fourth" are used in the claims, the description, and the drawings of the present application, they are only used to distinguish different objects and not to describe a specific order. The terms "comprising" and "including" used in the description and claims of the present application indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0090] It should also be understood that the terms used in the description of the present application herein are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the description and claims of the present application, unless the context clearly indicates otherwise, the singular forms "a", "an", and "the" are intended to include the plural forms. It should be further understood that the term "and / or" used in the description and claims of the present application refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0091] Although the embodiments of the present application are as above, the above content is only examples adopted for the convenience of understanding the present application, and is not intended to limit the scope and application scenarios of the present application. Any person skilled in the art within the technical field described in the present application can make any modifications and changes in the form of implementation and details without departing from the spirit and scope disclosed in the present application. However, the scope of patent protection of the present application shall still be subject to the scope defined by the appended claims.

Claims

1. A method for stitching a field of view in a charged particle beam system, characterized in that: The charged particle beam system comprises an optical navigation device, a magnetic deflection device and an electric deflection device, and the method comprises: Based on the first calibration, constructing a first relationship model between the optical navigation device and the magnetic deflection device at different working distances; Using a preprocessing model, the resolution of the first image taken by the optical navigation device is processed to match the second image taken by the magnetic deflection device; forming a first field of view switching point from the optical navigation device to the magnetic deflection device according to the first relationship model, the matched first image and the second image; constructing a second relationship model between the magnetic deflection device and the electric deflection device based on the second calibration; forming a second field of view switching point from the magnetic deflection device to the electric deflection device according to the second relationship model; Corresponding field of view compensation is performed based on the first field of view switching point and the second field of view switching point to achieve field of view stitching in a charged particle beam system.

2. The method according to claim 1, characterized in that The first calibration is performed by the following operations: moving a sample stage, wherein a first target and a second target are marked on the sample stage, and the first target is located at the center of the sample stage; Using the optical navigation device to take a plurality of first images of the sample stage to calibrate the distance relationship between the first images and the sample stage; Based on the distance relationship and using the magnetic deflection device, a second image of the sample stage at the target working distance is captured to calibrate the angular relationship between the first image and the second image at the target working distance.

3. The method according to claim 2, characterized in that The step of using the optical navigation device to capture a plurality of first images of the sample stage to calibrate the distance relationship between the first images and the sample stage includes: Acquire the actual position of the sample stage corresponding to each of the first images; Calculating movement distances of the first target and the second target on the corresponding first image; A distance matrix is ​​determined according to the actual position and the moving distance to calibrate the distance relationship between the first image and the sample stage.

4. The method according to claim 3, characterized in that Wherein, based on the distance relationship and using the magnetic deflection device to capture a second image of the sample stage at the target working distance, so as to calibrate the angular relationship between the first image and the second image at the target working distance comprises: Determine, based on the distance relationship, a first position where the first target and the second target are located on a first image at the target working distance; Acquire a second position of the first target and the second target on the second image at the target working distance; The angular relationship between the first image and the second image at a target working distance is calibrated according to the first position and the second position.

5. The method according to claim 2, characterized in that: Wherein, based on the first calibration, constructing a first relationship model between the optical navigation device and the magnetic deflection device at different working distances includes: Moving the sample stage, and calculating the angle deviation of the second image captured by the magnetic deflection device at different working distances; The relative angle difference between the optical navigation device and the magnetic deflection device is determined based on the first calibrated angle relationship and the angle deviation to construct the first relationship model between the optical navigation device and the magnetic deflection device at different working distances.

6. The method according to claim 5, characterized in that Wherein moving the sample stage and calculating the angle deviation of the second image captured by the magnetic deflection device at different working distances comprises: Moving the sample stage, and calculating the image deviation of the second image captured by the magnetic deflection device at the target feature area before and after the working distance is changed; The deflector of the magnetic deflection device is adjusted according to the image deviation and a preset threshold value until the image deviation is less than the preset threshold value, and a corresponding deflection angle is obtained to calculate the angle deviation of the second image taken by the magnetic deflection device at different working distances.

7. The method according to claim 1, characterized in that The preprocessing model includes a downsampling module and an upsampling module, and using the preprocessing model to perform resolution processing on the first image captured by the optical navigation device includes: Using the downsampling module to perform a downsampling operation on the first image captured by the optical navigation device to obtain a downsampling feature; Performing an upsampling operation on the downsampled features using the upsampling module, and fusing the corresponding downsampled features in the upsampling operation to obtain upsampled features; and The final up-sampling feature is superimposed on the first image to obtain a target image, so as to perform resolution processing on the first image taken by the optical navigation device.

8. The method according to claim 7, characterized in that The preprocessing model is a Restormer model.

9. The method according to claim 1, characterized in that: Wherein, based on the second calibration, constructing a second relationship model between the magnetic deflection device and the electric deflection device comprises: Based on the second calibration, the magnification and the angle difference between the deflector of the magnetic deflection device and the deflector of the electric deflection device are calculated to construct a second relationship model between the magnetic deflection device and the electric deflection device.

10. The method according to claim 9, characterized in that Wherein, based on the second calibration, calculating the magnification and the angle difference between the deflector of the magnetic deflection device and the deflector of the electric deflection device comprises: Respectively identifying a first grid number and a second grid number corresponding to the calibration sample under the magnetic deflection device and the electric deflection device; adjusting the amplification gain multiple of the deflector of the magnetic deflection device so that the first grid number is consistent with the second grid number, so as to calculate the amplification multiple between the deflector of the magnetic deflection device and the deflector of the electric deflection device; An angular deviation between the grids corresponding to the magnetic deflection device and the electric deflection device at the magnification is identified to calculate an angular difference between the deflector of the magnetic deflection device and the deflector of the electric deflection device.

11. The method according to claim 1, characterized in that: It also includes: Setting a third field of view switching point and a fourth field of view switching point; Corresponding field of view compensation is performed based on the first field of view switching point, the second field of view switching point, the third field of view switching point, and the fourth field of view switching point to achieve field of view stitching in a charged particle beam system.

12. The method according to claim 11, characterized in that The third field of view switching point and the fourth field of view switching point are reverse field of view switching of the first field of view switching point and the second field of view switching point.

13. An electronic device, characterized in that: include: processor; as well as A memory storing program instructions for field of view stitching in a charged particle beam system, wherein when the program instructions are executed by the processor, the electronic device implements the method according to any one of claims 1-12.

14. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, on which computer-readable instructions for field of view stitching in a charged particle beam system are stored. When the computer-readable instructions are executed by one or more processors, the method described in any one of claims 1 to 12 is implemented.

Images