SMT-AOI AI re-judgment system data automatic labeling method and device

Abstract

The invention is suitable for the technical field of surface mounting, and discloses an SMT-AOI AI re-judgment system data automatic labeling method and device, and the method comprises the steps: collecting first sample data according to the input operation of a re-judgment person in a user interface of an SMT-AOI re-judgment system; based on the first sample data, determining a template of an input operation effective range; cleaning the first sample data based on a template of the effective range of the input operation to obtain second sample data; and according to the type of the defect label, labeling the second sample data to obtain labeled target data. According to the technical scheme, the annotation data for training the AI model in the SMT-AOI AI re-judgment system can be automatically generated, extra annotation personnel are not needed, and the method is not limited by suppliers of the SMT-AOI re-judgment system.

Description

Technical Field

[0001] The present invention relates to the field of surface mount technology, and in particular to a method and device for automatically labeling data of an SMT-AOI AI review system. Background Art

[0002] Computer vision systems are currently being widely used in industrial visual inspection, robot vision guidance and automated assembly. In modern mass production, product quality inspection and quality control in process links are indispensable, such as parts appearance, product packaging, circuit board welding quality, etc.

[0003] SMT (Surface Mount Technology) is a key technology in the modern electronics manufacturing industry. It achieves circuit connection by mounting pinless or short-lead surface mount components (SMC / SMD for short, called sheet components in Chinese) on the surface of the printed circuit board (PCB) or other substrates, and soldering and assembling them through reflow soldering and other methods. However, due to quality problems in the production process, STM-AOI is introduced in the SMT production line to detect various defects generated in the SMT (surface mount technology) process in the electronics manufacturing industry. AOI, the full name of Automatic Optical Inspection, is an intelligent detection technology based on optical principles, image comparison principles and statistical modeling principles, which uses machine vision to replace manual visual inspection. It is widely used in the electronics manufacturing industry, especially in PCB (printed circuit board). The configuration of a typical SMT production line is as follows: Figure 1 shown.

[0004] Traditional STM-AOI equipment may misjudge or miss complex or subtle defects. Each SMT production line needs to be equipped with personnel to manually inspect the uncertainties detected by the AOI system and re-judge the results of the AOI inspection. During the normal production process of the SMT production line, manual re-judgment takes up a lot of human resources, which has the disadvantages of low work efficiency and high labor costs. Summary of the invention

[0005] To solve the above problems, the present invention provides an SMT-AOI AI re-judgment system data automatic labeling method and device, which can automatically generate labeling data for training the AI ​​model in the SMT-AOI AI re-judgment system without the need for additional labeling personnel and is not restricted by the SMT-AOI re-judgment system supplier.

[0006] To achieve the above objectives, the present disclosure adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a method for automatically labeling data in an SMT-AOI AI review system, comprising:

[0008] Collecting first sample data according to input operations of the review personnel in the user interface of the SMT-AOI review system;

[0009] Determine a template for a valid range of input operations based on the first sample data;

[0010] Cleaning the first sample data based on the template of the valid range of the input operation to obtain the second sample data;

[0011] The second sample data is labeled according to the type of the defect label to obtain labeled target data.

[0012] Further, according to the input operation of the review personnel in the user interface of the SMT-AOI review system, the first sample data is collected, including:

[0013] According to the input operation of the reviewer in the user interface of the SMT-AOI review system, a screenshot of the user interface and the position coordinates of the input operation are obtained;

[0014] Extract valid images from screenshots of the user interface;

[0015] Convert the position coordinates of the input operation into relative coordinates;

[0016] The valid image and relative coordinates are determined as the first sample data.

[0017] Further, based on the first sample data, determining a template for a valid range of an input operation includes:

[0018] Convert the valid image in the first sample data into a grayscale image;

[0019] Determine the effective pixel range of the interaction unit corresponding to the defect label based on the pixel position of the interaction unit corresponding to the defect label in the grayscale image data;

[0020] Based on the valid pixel range and the type of defect label, a template of the valid range of the input operation is determined.

[0021] Furthermore, based on the template of the valid range of the input operation, the first sample data is cleaned to obtain the second sample data, including:

[0022] Compare the relative coordinates of the input operation position in the first sample data with the template of the input operation valid range, and retain the relative coordinates of the input operation position if they are within the input operation valid range;

[0023] Based on the corresponding defect label type in the template of the valid range of the input operation, the valid images in the retained data are labeled with the type;

[0024] The retained data is determined as second sample data.

[0025] In a second aspect, the present invention further provides an SMT-AOI AI review system data automatic labeling device, comprising:

[0026] A collection module, used for collecting first sample data according to input operations of the review personnel in the user interface of the SMT-AOI review system;

[0027] A determination module, used to determine a template of a valid range of an input operation based on the first sample data;

[0028] A cleaning module, used for cleaning the first sample data based on the template of the valid range of the input operation to obtain the second sample data;

[0029] The labeling module is used to label the second sample data according to the type of the defect label to obtain the labeled target data.

[0030] Furthermore, the acquisition module is also used for:

[0031] According to the input operation of the reviewer in the user interface of the SMT-AOI review system, a screenshot of the user interface and the position coordinates of the input operation are obtained;

[0032] Extract valid images from screenshots of the user interface;

[0033] Convert the position coordinates of the input operation into relative coordinates;

[0034] The valid image and relative coordinates are determined as the first sample data.

[0035] Furthermore, the module is determined to be further used for:

[0036] Convert the valid image in the first sample data into a grayscale image;

[0037] Determine the effective pixel range of the interaction unit corresponding to the defect label based on the pixel position of the interaction unit corresponding to the defect label in the grayscale image data;

[0038] Based on the valid pixel range and the type of defect label, a template of the valid range of the input operation is determined.

[0039] Furthermore, the annotation module is also used to:

[0040] Compare the relative coordinates of the input operation position in the first sample data with the template of the input operation valid range, and retain the relative coordinates of the input operation position if they are within the input operation valid range;

[0041] Based on the corresponding defect label type in the template of the valid range of the input operation, the valid images in the retained data are labeled with the type;

[0042] The retained data is determined as second sample data.

[0043] In a third aspect, the present invention further provides an electronic device, comprising: a processor and a memory;

[0044] The processor is coupled to the memory;

[0045] The processor is used to read and execute the program or instructions stored in the memory, so that the device executes the method of the first aspect.

[0046] In a fourth aspect, the present invention further provides a computer-readable storage medium storing a computer program, which implements the method of the first aspect when executed by a processor.

[0047] In summary, the technical solution provided by this application has at least the following technical effects or advantages:

[0048] The technical solution of the present invention automatically generates the annotation data for training the AI ​​model in the SMT-AOI AI re-judgment system according to the operations of the re-judgment personnel in the user interface of the SMT-AOI re-judgment system during the normal production process, without the need for additional annotation personnel. Since the suppliers of the SMT-AOI re-judgment system are different and the SMT-AOI re-judgment system is a closed system, it is difficult to obtain data for operation through the API interface of the SMT-AOI re-judgment system. The technical solution of the present invention generates the annotation data of the AI ​​model based on the screenshot of the screen, and is not restricted by the supplier of the SMT-AOI re-judgment system.

[0049] Other features and advantages of the present disclosure will be described in the following description, and partly become apparent from the description, or be understood by implementing the present disclosure. The purpose and other advantages of the present disclosure can be realized and obtained by the structures pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0051] Figure 1 This is a schematic diagram of the configuration structure of a typical SMT production line;

[0052] Figure 2 A schematic diagram of a user interface of an SMT-AOI review system in an embodiment of the present invention;

[0053] Figure 3 It is a flow chart of a method for automatic data labeling of an SMT-AOI AI review system in an embodiment of the present invention;

[0054] Figure 4 A schematic diagram of a "black" screenshot of the user interface of the SMT-AOI review system in an embodiment of the present invention;

[0055] Figure 5 A schematic diagram of a grayscale image effective image range corresponding to a screenshot of a user interface of an SMT-AOI review system in an embodiment of the present invention;

[0056] Figure 6 It is a structural schematic diagram of an automatic data labeling device for an SMT-AOI AI review system provided in an embodiment of the present invention;

[0057] Figure 7 The present invention is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0058] In order to make the purpose, technical solution and advantages of the embodiments of the present disclosure clearer, the technical solution in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present disclosure.

[0059] In the SMT production line, STM-AOI is used to detect various defects generated in the normal production process of the SMT production line. The types of defects that SMT-AOI can detect include but are not limited to:

[0060] Component defects: component deviation, component lack of tin, solder joint short circuit, solder joint cold soldering, component reverse, component wrong installation, component deformation, component missing installation, component erection, etc.

[0061] Defects after printing: collapse, bridging, no solder paste, too little solder paste, too much solder paste, etc.

[0062] Defects after mounting: offset, missing components, sideways mounting, reverse polarity of components, etc.

[0063] Defects after welding: misalignment, bridging, tombstone, solder joints that are too small, solder joints that are too large, etc.

[0064] Other specific defects: pull tip, blow hole, solder bead, too little tin / too much tin on pad, bridge, lifted foot, gold finger contamination / scratches, etc.

[0065] For the images of the potential defective locations of electronic components detected by the SMT-AOI equipment, the review personnel are required to conduct manual review through the SMT-AOI review system. The user interface of the SMT-AOI review system is as follows: Figure 2 The reviewer compares the defect image with the reference image, performs input operations (such as clicking or touching) corresponding to the defect label in the user interface, and reviews the defect image.

[0066] With the gradual maturity of AI technology, integrating AI models into existing SMT production lines and training the AI ​​models, using trained AI models to replace some manual re-judgment operations can reduce the labor intensity of re-judgment personnel, improve work efficiency, and reduce costs. However, training AI models requires large-scale annotation data, and how to obtain annotation data has become an urgent problem to be solved. The above-mentioned system for re-judgment through AI models is the SMT-AOI AI re-judgment system. The technical problem to be solved by the technical solution of the present invention is how to obtain the annotation data required for training the SMT-AOI AI re-judgment system.

[0067] The technical solution of the present invention automatically generates the annotation data for training the AI ​​model in the SMT-AOI AI re-judgment system according to the operations of the re-judgment personnel in the user interface of the SMT-AOI re-judgment system during the normal production process, without the need for additional annotation personnel. Since the suppliers of the SMT-AOI re-judgment system are different and the SMT-AOI re-judgment system is a closed system, it is difficult to obtain data for operation through the API interface of the SMT-AOI re-judgment system. The technical solution of the present invention generates the annotation data of the AI ​​model based on the screenshot of the screen, and is not restricted by the supplier of the SMT-AOI re-judgment system.

[0068] Figure 3 Schematic diagram of an automatic data labeling method of an SMT-AOI AI review system in an embodiment of the present invention. As shown in the figure, the method includes:

[0069] S101, collecting first sample data according to the input operation of the review personnel in the user interface of the SMT-AOI review system;

[0070] Exemplarily, the SMT-AOI review system is connected to a display screen, which displays the user interface of the SMT-AOI review system, monitors the input operations of the review personnel in the user interface of the SMT-AOI review system in real time, takes a screenshot of the current user interface according to the input operation, and records the position coordinates (x, y) of the input operation, which grows to the right and downward from the origin (0, 0) in the upper left corner of the display screen. That is, when the input operation is performed, the pixel position of the input position relative to the upper left corner of the screen is recorded. The input operation can be a mouse click input, touch input, or other interactive behavior operations. A valid image is captured from the screenshot of the current user interface, and the position coordinates of the input operation are converted into relative coordinates. The valid image and the relative coordinates are the first sample data. As needed, based on the time of the input operation, the valid image of the screenshot and the relative coordinates of the input operation can be stored as an image file and a text file, respectively.

[0071] For example, in real-time monitoring of the mouse click operation of the review personnel in the user interface of the SMT-AOI review system, when the left button is clicked, the pixel position (x, y) of the click position relative to the upper left corner of the screen is recorded, that is, the absolute coordinates. The value of (x, y) is (317, 76), 317 represents the horizontal coordinate of the click position (the pixel distance from left to right), and 76 represents the vertical coordinate of the click position (the pixel distance from top to bottom).

[0072] The SMT-AOI review system runs on the Windows platform, and takes a screenshot of the current user interface of the SMT-AOI review system based on the API standard of the Windows platform.

[0073] The implementation process of screenshot technology on Windows platform is as follows:

[0074] (1) Get device context: Use the Windows API function GetDC to get the screen device context. The device context is a structure that contains device drawing information.

[0075] (2) Create a compatible memory device context: Use CreateCompetibleDC to create a memory device context that is compatible with the screen device context.

[0076] (3) Create a compatible bitmap: Use CreateCompatibleBitmap to create a bitmap that is compatible with the screen device context. This bitmap will be used to store the screenshot.

[0077] (4) Select the bitmap into the memory device context: Use the SelectObject function to select the created bitmap into the memory device context.

[0078] (5) Bit block transfer (BitBlt): Use the BitBlt function to transfer the contents in the screen device context to the bitmap in the memory device context.

[0079] (6) Access bitmap data: Use the GetDIBits function to obtain the byte data of the bitmap.

[0080] (7) Clean up resources: release device context and bitmap to avoid resource leakage.

[0081] The above process is implemented based on the API in the grabscreen_win32 standard library of C and C++. The byte data of the above bitmap is the screenshot image data of the current user interface of the SMT-AOI review system when the input operation is monitored.

[0082] In actual use, due to different monitors, different Windows operating versions, or monitor settings, the use of grabscreen_win32 may cause "black" problems (such as Figure 4 Therefore, it is necessary to determine whether there is "black" in the screenshot image and remove the "black" part.

[0083] The steps to determine the "black" and remove the "black" part are as follows:

[0084] Convert the screenshot image into a grayscale image, calculate the valid image range, and capture the valid image.

[0085] Since the screenshot is a color image with three color channels, it needs to be converted into a grayscale image for calculation convenience.

[0086] Since the "black" in the screenshot is not the real display image of the current user interface on the screen, the pixel values ​​of the "black" part of the pure black image are all 0. The valid range of the screenshot can be confirmed by judging that the pixel values ​​after a certain row and a certain column in the grayscale image are all "0" (such as Figure 5 According to the effective range of the screenshot image, the image data is intercepted to obtain a valid image of the screenshot image.

[0087] Due to factors such as different screens and different resolutions of display screens, the absolute coordinates of the click position are different. However, the relative positions of the layouts in the user interface of the same version of the SMT-AOI review system from the same supplier are fixed. Therefore, the coordinates of the click position need to be normalized. Based on the effective image of the screenshot, the click position coordinates are converted into the relative position (normalized_x, normalized_x) between (0,1). The conversion method is:

[0088] normalized_x = x / screen_width

[0089] normalized_y=y / screen_height

[0090] Among them, screen_width represents the width of the effective screenshot image, screen_height represents the height of the effective screenshot image, and x, y represent the absolute coordinates of the click position. Assume screen_width = 1919, screen_height = 1079, x∈(0,1919)y∈(0,1079), the absolute coordinates of the click position (x, y) = (317, 76) are converted into relative coordinates (normalized_x, mornalized_y) = (317 / 1919, 76 / 1079).

[0091] The relative coordinates of the effective image and click position of the above screenshot image are determined as the first sample data

[0092] The effective image of the above screenshot can be stored as an image file as needed, and the relative coordinates of the click position can be stored as a text file. The time of the click operation is used as the identifier of the image file and the text file name.

[0093] S102: Determine a template of a valid range of input operation based on the first sample data;

[0094] Exemplarily, the relative positions of the layouts in the user interface of the same version of the SMT-AOI review system of the same supplier are fixed. The user interface includes: a defect picture display module, a reference picture display module, and an input operation interaction module. The defect picture display module is used to display pictures of potential locations of defects in electronic components detected by the STM-AOI device; the reference picture module is used to display reference pictures to facilitate the review personnel to compare the defect pictures with the reference pictures; the input operation interaction module is used to interact with the input operations corresponding to the defect labels, and each defect label corresponds to an interaction unit (such as a button, touch block). Defect labels include: short circuit, component NG, body warping, pin warping, missing parts, less tin, unmelted tin, polarity, flipping, offset, solder, exposed copper, wrong parts, misjudgment, and defective. Defect labels also include the various defects generated by the above-mentioned SMT production line detected by STM-AOI during normal production.

[0095] Based on the valid image in the first sample data, it is converted into a grayscale image. In the grayscale image data, the pixels of the interaction unit corresponding to each defect label have relatively obvious features. The pixel positions describing each interaction unit are extracted according to the features, that is, the absolute coordinates, which are normalized to obtain the corresponding relative coordinates, that is, the effective pixel positions of each interaction unit. According to the effective pixel positions of each interaction unit, the effective pixel range of each interaction unit is determined. Based on the defect label and the effective pixel range of each interaction unit, the template of the effective range of the input operation is determined (as shown in Table 1).

[0096] Table 1

[0097] Label X_Start X_End Y_Start Y_End 01 0.403856 0.464304 0.784986 0.836886 02 0.464311 0.524759 0.784986 0.836886

[0098] In Table 1, Label indicates the type of defect label, X_Start, X_End, X_End, and Y_End represent the effective pixel range of the interaction unit corresponding to the defect label, X_Start indicates the start position in the horizontal direction, X_End indicates the end position in the horizontal direction, Y_Start indicates the start position in the vertical direction, and Y_End indicates the end position in the vertical direction. Table 1 only shows some sample data, and more complete data can be included according to actual conditions.

[0099] S103, based on the template of the input operation valid range, clean the first sample data to obtain second sample data;

[0100] Exemplarily, the relative coordinates of the input operation position in the first sample data are compared with the template of the valid range of the input operation. If the relative coordinates of the input operation position are within the valid range of the input operation, the data is valid data and is retained. At the same time, the data is labeled according to the corresponding defect label type in the template of the valid range of the input operation; if the relative coordinates of the input operation position are not within the valid range of the input operation, the data is discarded or deleted.

[0101] According to the above comparison operation, the data retained in the first sample data is determined as the second sample data;

[0102] S104: Label the second sample data according to the type of the defect label to obtain labeled target data.

[0103] Exemplarily, the relative positions of the layouts in the user interface of the same version of the SMT-AOI review system of the same supplier are fixed. The valid image in the second sample data, i.e., the valid image of the screenshot image of the user interface of the SMT-AOI review system, and the defect image display module in the user interface corresponding to the valid image display at least one defect image. According to the coordinate value corresponding to the fixed position of the defect image in the user interface, the corresponding annotation image can be intercepted from the valid image.

[0104] If the defect image display module displays multiple defect images, the number of the intercepted annotated image is determined according to the layout positions of the multiple defect images in the user interface. The number can be a number or other ordered characters.

[0105] The annotation information is constructed according to the time of the input operation, the number of the annotated image, and the defect type corresponding to the valid image, such as "20240720_084334_XX_YY", where 20240720_084334 indicates the time of the input operation, XX indicates the number of the annotated image, and YY indicates the defect type corresponding to the valid image.

[0106] Based on the annotation information, the corresponding annotation image is annotated, that is, the annotation information is used as the file name, and the annotated image is saved as an image file.

[0107] The annotated image file is the target data, which is used to train the AI ​​model in the SMT-AOI AI review system.

[0108] Figure 6 It is a structural schematic diagram of an SMT-AOI AI review system data automatic labeling device provided in an embodiment of the present disclosure.

[0109] like Figure 6 As shown, the system includes: XX module, XX module.

[0110] It should be noted that, for the sake of convenience, Figure 6 For example, only the main modules of the XXX system structure are shown. In practical applications, the system may also include modules or components not shown in the figure; the system is not limited to the above module structure, and may also be other module structures that implement the above method embodiments.

[0111] Figure 7 A schematic diagram of the structure of an electronic device provided in an embodiment of the present disclosure.

[0112] like Figure 7 As shown, the electronic device includes: a processor and a memory;

[0113] The processor is used to read and execute programs and instructions stored in the memory, so that the electronic device executes the above method embodiment.

[0114] It should be noted that, for the sake of convenience, Figure 7 Only the main components of the electronic device are shown. In practical applications, the electronic device may also include components or assemblies not shown in the figure.

[0115] The present disclosure also provides a computer-readable storage medium storing a program or instruction. When a computer reads and executes the program or instruction, the computer executes the above method embodiment.

[0116] Although the present disclosure has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure.

Claims

1. A method for automatic data annotation of SMT-AOI AI re-judgment system, characterized in that: include: Collecting first sample data according to input operations of the review personnel in the user interface of the SMT-AOI review system; Determine a template for a valid range of input operations based on the first sample data; Based on the template of the valid range of the input operation, the first sample data is cleaned to obtain second sample data; The second sample data is labeled according to the type of the defect label to obtain labeled target data.

2. The automatic data labeling method of the SMT-AOI AI review system according to claim 1 is characterized in that: The collecting of the first sample data according to the input operation of the review personnel in the user interface of the SMT-AOI review system includes: According to the input operation of the reviewer in the user interface of the SMT-AOI review system, a screenshot of the user interface and the position coordinates of the input operation are obtained; Extracting valid images from the screenshots of the user interface; Converting the position coordinates of the input operation into relative coordinates; The valid picture and the relative coordinates are determined as first sample data.

3. The automatic data labeling method of the SMT-AOI AI review system according to claim 1 is characterized in that: The template for determining the valid range of the input operation based on the first sample data includes: Converting the valid image in the first sample data into a grayscale image; Determine the effective pixel range of the interaction unit corresponding to the defect label based on the pixel position of the interaction unit corresponding to the defect label in the grayscale image data; Based on the valid pixel range and the type of the defect label, a template of the input operation valid range is determined.

4. The automatic data labeling method of the SMT-AOI AI review system according to any one of claims 1 to 3, characterized in that: The step of cleaning the first sample data based on the template of the valid range of the input operation to obtain the second sample data includes: Comparing the relative coordinates of the input operation position in the first sample data with the template of the input operation valid range, and retaining the relative coordinates of the input operation position if they are within the input operation valid range; Based on the corresponding defect label type in the template of the valid range of the input operation, annotate the valid images in the retained data; The retained data is determined as second sample data.

5. An automatic data labeling device for SMT-AOI AI review system, characterized in that: include: A collection module, used for collecting first sample data according to input operations of the review personnel in the user interface of the SMT-AOI review system; A determination module, used for determining a template of a valid range of input operations based on the first sample data; a cleaning module, configured to clean the first sample data based on the template of the valid range of the input operation to obtain second sample data; The labeling module is used to label the second sample data according to the type of the defect label to obtain labeled target data.

6. The SMT-AOI AI review system data automatic labeling device according to claim 5 is characterized in that: The acquisition module is also used for: According to the input operation of the reviewer in the user interface of the SMT-AOI review system, a screenshot of the user interface and the position coordinates of the input operation are obtained; Extracting valid images from the screenshots of the user interface; Converting the position coordinates of the input operation into relative coordinates; The valid picture and the relative coordinates are determined as first sample data.

7. The SMT-AOI AI review system data automatic labeling device according to claim 5, characterized in that: The determining module is further used for: Converting the valid image in the first sample data into a grayscale image; Determine the effective pixel range of the interaction unit corresponding to the defect label based on the pixel position of the interaction unit corresponding to the defect label in the grayscale image data; Based on the valid pixel range and the type of the defect label, a template of the input operation valid range is determined.

8. The SMT-AOI AI review system data automatic labeling device according to any one of claims 5 to 7, characterized in that: The annotation module is also used for: Comparing the relative coordinates of the input operation position in the first sample data with the template of the input operation valid range, and retaining the relative coordinates of the input operation position if they are within the input operation valid range; Based on the corresponding defect label type in the template of the valid range of the input operation, annotate the valid images in the retained data; The retained data is determined as second sample data.

9. An electronic device, characterized in that: include: Processor and memory; The processor is coupled to the memory; The processor is used to read and execute the program or instruction stored in the memory, so that the device executes the method according to any one of claims 1 to 4.

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

Images