Picture classification method and device of graphical programming platform based on artificial intelligence

Abstract

The invention discloses a picture classification method and device of a graphical programming platform based on artificial intelligence. The method comprises the following steps: acquiring a picture data set uploaded by a user through a graphical programming platform, labeling the picture data set, and performing enhancement processing on the labeled picture data set; based on the marked features of the picture data set, recommending a matched deep learning model, and receiving an adjustment operation of a user on hyper-parameters of the deep learning model through the graphical programming platform; and training the deep learning model by using the marked picture data set, generating a loss function curve and an accuracy change chart of the deep learning model in a training process in real time, and feeding back the training progress of the deep learning model in a visual mode. The technical problems that the technical threshold is high and the children are difficult to master in the existing AI learning process of the children are solved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence, and in particular, to a method and device for image classification of a graphical programming platform based on artificial intelligence. Background Art

[0002] In recent years, with the rapid development of artificial intelligence technology, the importance of programming and AI education in the field of children's education has become increasingly prominent. Educational institutions have incorporated computational thinking and AI literacy into the basic education system, aiming to cultivate children's ability to adapt to the future intelligent society. However, traditional programming education is mainly targeted at adolescent or adult learners, and its teaching content and tool design often fail to meet the special learning needs of children.

[0003] In view of the above problems, no effective solution has been proposed yet. Summary of the Invention

[0004] Embodiments of the present invention provide a method and device for image classification of a graphical programming platform based on artificial intelligence, so as to at least solve the technical problems of high technical threshold and difficulty in getting started existing in the current children's AI learning process.

[0005] According to one aspect of the embodiments of the present invention, there is provided a method for image classification of a graphical programming platform based on artificial intelligence, including: obtaining a picture data set uploaded by a user through the graphical programming platform, annotating the picture data set, and performing enhancement processing on the annotated picture data set; recommending a matching deep learning model based on the features of the annotated picture data set, and receiving, through the graphical programming platform, an operation of a user to adjust hyperparameters of the deep learning model; training the deep learning model by using the annotated picture data set, and generating in real time a loss function curve and an accuracy change chart of the deep learning model during the training process, and feeding back the training progress of the deep learning model in a visual manner.

[0006] According to another aspect of the embodiments of the present invention, there is also provided a device for image classification of a graphical programming platform based on artificial intelligence, including: a processing module configured to obtain a picture data set uploaded by a user through the graphical programming platform, annotate the picture data set, and perform enhancement processing on the annotated picture data set; a matching module configured to recommend a matching deep learning model based on the features of the annotated picture data set, and receive, through the graphical programming platform, an operation of a user to adjust hyperparameters of the deep learning model; a training module configured to train the deep learning model by using the annotated picture data set, and generate in real time a loss function curve and an accuracy change chart of the deep learning model during the training process, and feed back the training progress of the deep learning model in a visual manner.

[0007] In an embodiment of the present invention, an image dataset uploaded by a user is obtained through a graphical programming platform, the image dataset is labeled, and the labeled image dataset is enhanced; based on the features of the labeled image dataset, a matching deep learning model is recommended, and the graphical programming platform receives an operation for adjusting hyperparameters of the deep learning model by the user; the labeled image dataset is used to train the deep learning model, and a loss function curve and an accuracy change chart of the deep learning model during the training process are generated in real time, and the training progress of the deep learning model is fed back in a visual manner. Through the above solution, the technical problems of high technical threshold and difficulty in getting started existing in the existing children's AI learning process are solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The drawings described herein are used to provide a further understanding of the present invention, and constitute a part of this application. The illustrative embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0009] Figure 1 is a flowchart of a method for image classification of a graphical programming platform based on artificial intelligence according to an embodiment of the present invention;

[0010] Figure 2 is a flowchart of another method for image classification based on a graphical programming platform according to an embodiment of the present invention;

[0011] Figure 3 is a flowchart of yet another method for image classification of a graphical programming platform based on artificial intelligence according to an embodiment of the present invention;

[0012] Figure 4 is a flowchart of automatically labeling a dataset according to an embodiment of the present invention;

[0013] Figure 5 is a flowchart of dynamically optimizing a labeling model through closed-loop feedback according to an embodiment of the present invention;

[0014] Figure 6 is a flowchart of a method for starting training according to an embodiment of the present invention;

[0015] Figure 7 is a schematic structural diagram of an image classification device of a graphical programming platform based on artificial intelligence according to an embodiment of the present invention;

[0016] Figure 8 shows a schematic structural diagram of an electronic device suitable for implementing embodiments of the present disclosure. DETAILED DESCRIPTION

[0017] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0018] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present invention described herein can be implemented in an order different from those illustrated or described herein. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0019] According to an embodiment of the present invention, a method embodiment of a picture classification method for a graphical programming platform based on artificial intelligence is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions. And although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0020] Currently, in the field of children's education, programming and AI education have gradually received attention. However, many traditional programming tools and learning platforms make it difficult for children to complete complex machine learning tasks without programming foundation. Therefore, the combination of graphical programming platforms and artificial intelligence technology has become a key innovation point for improving children's programming education.

[0021] Figure 1 is a picture classification method for a graphical programming platform based on artificial intelligence according to an embodiment of the present invention, as Figure 1 shown, the method includes the following steps:

[0022] Step S102, obtaining a picture data set uploaded by a user through a graphical programming platform, annotating the picture data set, and performing enhancement processing on the annotated picture data set;

[0023] First, annotate the dataset. Perform preliminary annotation on the picture dataset through a pre-trained automatic image recognition algorithm; receive the user's correction operations on the preliminary annotation through the visualization interface of the graphical programming platform, and generate the annotated picture dataset.

[0024] Next, perform enhancement processing. Perform rotation, brightness adjustment, and / or scaling operations on the picture dataset, and display the image changes before and after the enhancement processing through comparison graphs on the graphical programming platform.

[0025] Step S104: Recommend a matching deep learning model based on the features of the annotated picture dataset, and receive the user's adjustment operations on the hyperparameters of the deep learning model through the graphical programming platform.

[0026] For example, extract the feature dimensions, number of categories, and sample distribution of the annotated picture dataset; based on the feature dimensions, number of categories, and sample distribution, perform matching in a preset model library, and recommend the model with the highest matching degree as the deep learning model. After that, receive the user's adjustment operations on the hyperparameters of the deep learning model through the graphical programming platform.

[0027] Step S106: Train the deep learning model using the annotated picture dataset, and generate the loss function curve and accuracy change chart of the deep learning model during the training process in real time, so as to visually feedback the training progress of the deep learning model.

[0028] For example, adjust the learning rate, batch size, or number of training epochs through the slider or dropdown menu provided by the graphical programming platform; display the impact of parameter adjustment on the training results through a real-time updated visualization chart, where the impact on the training results includes the loss function curve and the accuracy change chart.

[0029] The present invention provides another picture classification method based on a graphical programming platform, especially applied in children's education. By graphically and visually presenting the picture classification task, children can complete model training, optimization, and testing through simple operations such as dragging and clicking without understanding programming, thereby cultivating their logical thinking ability, innovation awareness, and AI foundation. This method is as Figure 2 shown, and includes the following steps:

[0030] Step S202: Provide a graphical programming interface.

[0031] The graphical programming platform provides an intuitive graphical interface for users, allowing them to construct training processes through simple drag-and-drop operations. Each task or operation module (such as data input, model selection, training parameter adjustment, etc.) has a corresponding icon. Users can simply drag and connect these modules through the graphical interface to achieve functions such as data import, preprocessing, model training, and evaluation. This interface greatly reduces the technical threshold, enabling children to complete the entire process from data import to model training without having to master complex code.

[0032] Step S204, intelligent data annotation and processing.

[0033] To reduce the burden on children during data annotation, the present invention provides an intelligent annotation assistance function. The platform can help children quickly annotate image datasets through automatic image recognition algorithms, reducing manual intervention. In addition, the platform uses visualization tools to show children the data preprocessing and enhancement effects (such as image rotation, brightness adjustment, etc.), enabling them to more intuitively understand the role of data in training.

[0034] Step S206, automatic model recommendation and fine-tuning.

[0035] Based on the characteristics of the uploaded dataset, the platform automatically recommends suitable image classification models, and children can start training directly according to the recommended models. The platform also provides hyperparameter adjustment options (such as learning rate, batch size, etc.), and uses visualization to help children understand the role of each parameter, enhancing their understanding of basic machine learning concepts.

[0036] Step S208, real-time training feedback and visualization results.

[0037] During the image classification training process, the platform provides real-time feedback, intuitively showing the progress and effect of model training through graphical charts such as loss function curves and accuracy changes. Children can understand the status of model training through this real-time feedback, and thus know when to adjust training parameters or terminate training. The platform uses simple and easy-to-understand charts to enable children to understand and master the basic concepts of "training" and "optimization" in machine learning.

[0038] Step S210, task guidance and learning suggestions.

[0039] The platform is built with a learning guidance system that provides short tips and suggestions for each step and module, helping children understand the purpose of the current operation at each stage. For example, during training, the platform will prompt children about the basic principles of model optimization or provide suggestions for pausing and tuning during training. This intelligent guidance function helps children better master the AI training process and does not rely too much on adult guidance.

[0040] Step S212, Cross-platform deployment and sharing.

[0041] After training, the trained image classification model is quickly deployed to multiple platforms, such as the Web side, desktop applications, etc., and even supports sharing the model with other users for model sharing and communication. This rapid deployment and sharing function not only improves children's practical operation ability but also allows them to see the effects of the models they trained in actual applications.

[0042] Step S214, Give rewards.

[0043] To improve children's participation and learning interest, the present invention combines gamification elements and designs a "level" learning path and reward mechanism during the training process. For each task completed (such as uploading data, training the model, evaluating results, etc.), children can obtain corresponding points or badges. These reward mechanisms stimulate children's learning motivation and also improve their task completion rate and academic performance.

[0044] In addition, elements of social learning are introduced. Children can interact with their peers or teachers, share the results of their models, discuss different training methods, and complete common tasks through group cooperation. This social function not only cultivates children's teamwork spirit but also enhances their communication and expression abilities.

[0045] The present invention not only solves the problem of technical usability but also reconstructs the AI education paradigm, transforms abstract machine learning concepts into tangible graphic modules (such as representing convolutional layers with building blocks), establishes a cognitive closed-loop of "operation-effect" through real-time feedback, and forms a sustainable learning community by using social interactions.

[0046] Figure 3 It is a method for classifying pictures of a graphical programming platform based on artificial intelligence according to another embodiment of the present invention. As Figure 3 shown, the method includes the following steps:

[0047] Step S302, The user imports a picture data set through a graphical interface and performs preliminary preprocessing.

[0048] First, the graphical programming platform uploads the picture data set stored locally or in the cloud by dragging the "data import" module in the graphical interface. The platform automatically validates the format of the uploaded pictures and only supports common image formats such as JPEG and PNG. For pictures with incompatible formats, the system triggers the conversion module to uniformly convert them to a preset format (such as RGB format with 224×224 pixels).

[0049] Next, the platform preprocesses and enhances the dataset. For example, all images are resized to a unified size (such as 224×224) through the bilinear interpolation algorithm to eliminate the impact of size differences on model training; the median filter is used to denoise low-quality images (such as blurred or noisy images); the EXIF information of the images (such as shooting time, device model) is parsed and stored in the database for reference in the subsequent annotation process. After preprocessing, rotation, brightness adjustment, and / or scaling operations can also be performed on the preprocessed image dataset, and the image changes before and after enhancement processing are displayed on the graphical programming platform through comparison graphs.

[0050] Step S304, generate initial labels based on the automatic annotation algorithm.

[0051] The platform calls a pre-trained lightweight image classification model (such as MobileNetV3) to automatically annotate the dataset. The specific process is as Figure 4 shown and includes the following steps:

[0052] Step S3042, feature extraction and inference.

[0053] The preprocessed and enhanced images are input into the model to extract high-dimensional feature vectors, and the class probability distribution is output through the fully connected layer.

[0054] Step S3044, generate initial labels.

[0055] Select the class with the highest probability as the initial label. If the highest probability is lower than the preset threshold, it is marked as "unclassified".

[0056] Step S3046, manual annotation.

[0057] Users can view the initial labels of each image through the graphical interface and modify them. For example, children can drag the "label correction" module to reclassify the images mislabeled as "dog" instead of "cat".

[0058] The present invention introduces a real-time feedback mechanism. The system records the user's modification behavior of the initial labels, including the number of modifications, the label differences before and after modification, the modification time interval, etc., to form a user behavior log; in addition, every N modifications (such as every 10 times), the system starts an online learning process, uses the labels modified by the user as new training data, and updates the weights of the last fully connected layer of the annotation model through an incremental learning algorithm (such as online stochastic gradient descent); for frequently modified classes (such as users repeatedly correcting "bird" to "airplane"), the system dynamically increases the weight of this class in the loss function to strengthen the feature learning of the model for this class. For example, when a child continuously modifies 3 "bird" images to "airplane", the platform automatically adjusts the annotation model to preferentially match the features of the "airplane" class in subsequent annotations and reduce similar errors.

[0059] Step S306: Based on the user behavior log, optimize the annotation model in real time and generate an enhanced annotation dataset.

[0060] The dynamic optimization of the annotation model is achieved through closed-loop feedback. As Figure 5 shown, the dynamic optimization method includes the following steps:

[0061] Step S3062: Data augmentation and feature fusion.

[0062] For the pictures modified by the user, the system automatically applies the data augmentation strategy to generate enhanced multi-version pictures; the feature vectors of the original pictures and the enhanced pictures are weighted and fused to improve the robustness of the model to perspective and illumination changes.

[0063] Step S3064: Online learning and model fine-tuning.

[0064] Adopt a lightweight online learning framework, only update the parameters of the last layer of the annotation model, and avoid the computational overhead caused by global retraining; design the loss function as the weighted sum of the cross-entropy loss and the user correction consistency:

[0065] L = α·L CE +(1 - α)·L Consistency

[0066] where L Consistency is calculated by comparing the label distribution differences before and after the user's modification to ensure the consistency between the model prediction and the user's intention. α is the weight coefficient, LCE is the cross-entropy loss, L is the loss function, and L Figure 1 is the consistency loss, which is calculated based on the label distribution qi after the user's correction, the model prediction distribution pi, and the KL divergence D Consistency KL KL edit

[0067]

[0068] where N represents the number of samples modified by the user.

[0069] Step S3066: Annotation quality assessment.

[0070] The system calculates the Annotation Consistency Index (ACI), which is defined as the inverse ratio of the number of user modifications to the automatic annotation error rate:

[0071]

[0072] where N edit is the number of user modifications, and Error Rate is the error rate.

[0073] If the ACI is lower than the preset threshold, the platform triggers a secondary annotation process to prompt the user to re-review the disputed samples. The finally generated annotated dataset includes the original images, enhanced images, and their optimized labels for subsequent model training.

[0074] Step S308: Train an image classification model based on the enhanced annotated dataset and provide real-time feedback and parameter tuning suggestions.

[0075] The user starts the training process by dragging the "Model Training" module. As Figure 6 shown, it includes the following steps:

[0076] Step S3082: Recommend an automated model.

[0077] The platform selects a pre-trained model based on the dataset size (e.g., recommend MobileNet for less than 1000 images and ResNet-18 for more than 1000 images) and the number of classes (e.g., recommend EfficientNet for multi-classes); the recommendation strategy is based on a predefined rule engine and dynamically matches the best model in combination with dataset features (such as resolution, class balance).

[0078] Step S3084: Visualize the training process.

[0079] The loss function curve, accuracy change, and confusion matrix are displayed in real time. The loss curve is additionally marked with the "User Correction Impact Area" to reflect the contribution of the user's modification behavior in step S306 to model convergence; for example, when the accuracy of the "airplane" class suddenly increases, the system prompts "This improvement may be related to your recent correction of the 'bird' class label."

[0080] Step S3086: Adjust interactive parameters.

[0081] The user can adjust parameters such as the learning rate (e.g., 0.001 - 0.1) and batch size (e.g., 16 - 128) through sliders, and the system provides intelligent suggestions based on the current training status (such as "Overfitting detected, it is recommended to reduce the learning rate by 30%"); the parameter adjustment strategy is based on a reinforcement learning algorithm, and balances model performance and training speed through an exploration-exploitation mechanism.

[0082] Step S310: Deploy the trained model to multiple platforms and support social sharing and collaborative optimization.

[0083] After training, the platform converts the model into ONNX format, which supports deployment to the Web (TensorFlow.js), mobile (Core ML) or embedded devices (TensorFlow Lite). After deployment, users can upload test images to view the classification results in real time, and the system will visually display the decision basis of the model through the highlighted activation area (Grad-CAM). Users can share the model to the community, and other users can test it and submit feedback.

[0084] The present invention combines a graphical programming platform with artificial intelligence technology, so that children can complete the entire process of image classification tasks directly through simple operations such as dragging and clicking without relying on traditional programming. Through intelligent training guidance, real-time feedback, cross-platform deployment and other functions, children can understand and master basic machine learning concepts and improve their programming and logical thinking abilities. The platform not only makes AI education more popular, but also makes the learning process of AI technology more interesting and challenging.

[0085] The embodiments of the present invention have the following beneficial effects: 1) Lowering the technical threshold: The graphical programming platform enables children to complete image classification tasks without having to understand complex programming languages, thus lowering the learning threshold. 2) Intelligent guidance: Intelligent data labeling, model recommendation, and hyperparameter fine-tuning functions help children get started quickly and obtain immediate feedback, thereby improving learning outcomes. 3) Gamification and reward mechanisms: Gamified learning paths and reward mechanisms increase children’s motivation and interest in learning. 4) Socialized learning: Social and interactive functions enhance children’s teamwork skills and promote their communication and learning with others.

[0086] The graphical programming platform of the present invention not only makes the training of image classification tasks easy and understandable, but also has broad application prospects in children's education through innovative educational models. By combining artificial intelligence with children's education, the platform can cultivate children's machine learning and AI foundation, stimulate their interest, and provide new educational tools for future AI talent reserves.

[0087] The present application also provides an image classification device based on an artificial intelligence graphical programming platform, such as Figure 7As shown in the figure, it includes: a processing module 72, configured to obtain the picture data set uploaded by the user through a graphical programming platform, annotate the picture data set, and perform enhancement processing on the annotated picture data set; a matching module 74, configured to recommend a matching deep learning model based on the features of the annotated picture data set, and receive the user's adjustment operation on the hyperparameters of the deep learning model through the graphical programming platform; a training module 76, configured to train the deep learning model using the annotated picture data set, and generate in real time the loss function curve and the accuracy change chart of the deep learning model during the training process, and feedback the training progress of the deep learning model in a visual manner.

[0088] It should be noted that: the picture classification device of the graphical programming platform based on artificial intelligence provided in the above embodiment is only illustrated by the division of the above functional modules. In practical applications, the above functions can be allocated to different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the picture classification device of the graphical programming platform based on artificial intelligence provided in the above embodiment and the method embodiment of the picture classification method of the graphical programming platform based on artificial intelligence belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0089] Figure 8 The structure diagram of the electronic device suitable for implementing the embodiments of the present disclosure is shown. It should be noted that Figure 8 The electronic device shown is only an example and should not bring any limitation to the functions and usage scope of the embodiments of the present disclosure.

[0090] As Figure 8 shown, the electronic device includes a central processing unit (CPU) 1001, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 1002 or the program loaded from the storage part 1008 into the random access memory (RAM) 1003. In the RAM 1003, various programs and data required for system operation are also stored. The CPU 1001, ROM 1002, and RAM 1003 are connected to each other through a bus 1004. The input / output (I / O) interface 1005 is also connected to the bus 1004.

[0091] The following components are connected to the I / O interface 1005: an input section 1006 including a keyboard, a mouse, etc.; an output section 1007 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc. and a speaker, etc.; a storage section 1008 including a hard disk, etc.; and a communication section 1009 including a network interface card such as a LAN card, a modem, etc. The communication section 1009 performs communication processing via a network such as the Internet. A drive 1010 is also connected to the I / O interface 1005 as required. A removable medium 1011 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc. is mounted on the drive 1010 as required so that a computer program read therefrom is installed in the storage section 1008 as required.

[0092] The above are only the preferred embodiments of the present application. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present application, several improvements and modifications can be made, and these improvements and modifications should also be regarded as the protection scope of the present application.

Claims

1. A method for image classification of a graphical programming platform based on artificial intelligence, characterized in that, Including: Obtain the image dataset uploaded by the user through a graphical programming platform, annotate the image dataset, and perform enhancement processing on the annotated image dataset; Based on the characteristics of the annotated image dataset, recommend a matching deep learning model, and receive the user's adjustment operation on the hyperparameters of the deep learning model through the graphical programming platform; Use the annotated image dataset to train the deep learning model, and generate the loss function curve and accuracy change chart of the deep learning model during the training process in real time, and feedback the training progress of the deep learning model in a visual manner.

2. The method according to claim 1, wherein Annotating the image dataset includes: Perform preliminary annotation on the image dataset through a pre-trained automatic image recognition algorithm; Receive the user's correction operation on the preliminary annotation through the visual interface of the graphical programming platform, and generate the annotated image dataset.

3. The method according to claim 1, wherein Performing enhancement processing on the annotated image dataset includes: performing rotation, brightness adjustment, and / or scaling operations on the image dataset, and displaying the image changes before and after the enhancement processing through a comparison chart on the graphical programming platform.

4. The method according to claim 3, characterized in that Recommending a matching deep learning model based on the characteristics of the annotated image dataset includes: Extract the feature dimension, number of categories, and sample distribution of the annotated image dataset; Based on the feature dimension, number of categories, and sample distribution, perform matching in a preset model library, and recommend the model with the highest matching degree as the deep learning model.

5. The method according to claim 1, wherein Receiving the user's adjustment operation on the hyperparameters of the deep learning model through the graphical programming platform includes: adjusting the learning rate, batch size, or number of training epochs through the slider or dropdown menu provided by the graphical programming platform; Generating the loss function curve and accuracy change chart of the deep learning model during the training process in real time includes: displaying the impact of parameter adjustment on the training result through a real-time updated visual chart, wherein the impact on the training result includes the loss function curve and the accuracy change chart.

6. The method according to any one of claims 1 to 5, characterized in that, After training the deep learning model, the method further includes: converting the trained deep learning model into a lightweight format, and integrating it into the graphical programming platform as a building block for use when the user performs graphical programming to classify images.

7. An image classification device for a graphical programming platform based on artificial intelligence, characterized in that, Including: A processing module configured to obtain the image dataset uploaded by the user through a graphical programming platform, annotate the image dataset, and perform enhancement processing on the annotated image dataset; A matching module configured to recommend a matching deep learning model based on the characteristics of the annotated image dataset, and receive the user's adjustment operation on the hyperparameters of the deep learning model through the graphical programming platform; A training module, configured to train the deep learning model by using the labeled image dataset, and generate in real time a loss function curve and an accuracy change chart of the deep learning model during the training process, and feed back the training progress of the deep learning model in a visual manner.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein, when the program runs, it controls the device where the computer-readable storage medium is located to execute the method according to any one of claims 1 to 6.

9. A computer device, characterized in that, Comprising: A memory and a processor, The memory stores a computer program; The processor is configured to execute the computer program stored in the memory, and when the computer program runs, the processor executes the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images