A grouping whiteboard control method for automatic split-screen display

Abstract

The application provides a grouping whiteboard control method for automatic split-screen display, relates to the technical field of split-screen display, and dynamically matches adaptive resources for each group by collecting multi-dimensional information of users and analyzing data characteristics; a corresponding number of split-screen windows are automatically generated according to the number of groups based on a layout grid algorithm, and an association relationship between the windows and the groups is established; resources are displayed in corresponding windows in an intuitive and efficient form by means of a data visualization model, user operations are detected in real time, multi-group synchronous operations are realized, and real-time information sharing is ensured; the identity of a member is recognized, the concentration degree and understanding state of the member on current teaching resources are judged, an adjustment strategy of the split-screen window is generated in combination with the difficulty level of the teaching resources, and the size and position of the split-screen window are dynamically adjusted based on the adjustment strategy, so that the smoothness and visualization efficiency of group cooperation are significantly improved, and a more efficient and intelligent interactive experience is brought to teaching and team cooperation.

Description

Technical Field

[0001] This invention relates to the field of split-screen display, and more particularly to a method for controlling a grouped whiteboard for automatic split-screen display. Background Technology

[0002] Currently, electronic whiteboards are mostly used in educational and training settings, facilitating the intuitive display of content in a multimedia format.

[0003] In a teaching setting, teachers use interactive whiteboards to display teaching content to students. However, this approach has significant limitations in multi-group collaborative scenarios. Because different groups vary greatly in terms of learning ability, pace, and areas of weakness, their learning plans differ, resulting in different teaching resources allocated to each group and consequently, different content displayed on the interactive whiteboard.

[0004] To enable different content to be displayed on the electronic whiteboard for different groups, the traditional method is to use screen splitting on the electronic whiteboard. However, when splitting the screen, it is necessary to divide the screen area in advance to determine the size and position of the split window, and manually set the number of split windows. This method is difficult to adapt to collaborative scenarios of different group sizes and needs. If the number of groups changes, the screen splitting needs to be manually adjusted, which is inefficient. In addition, displaying content of different difficulty levels with the same window size results in a waste of resources.

[0005] Therefore, it is crucial to design a method that can automatically generate a corresponding number and appropriate size of split-screen windows on the electronic whiteboard, and dynamically adjust the position and size of the split-screen windows according to the displayed content. Summary of the Invention

[0006] To address the aforementioned issues, this invention provides a grouped whiteboard control method for automatic split-screen display. This method can automatically generate a corresponding number of split-screen windows of suitable size on the electronic whiteboard, and can dynamically adjust the position and size of the split-screen windows according to the displayed content, clearly and intuitively displaying resources in the corresponding windows.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A method for controlling a grouped whiteboard with automatic split-screen display includes the following steps: S1. Collect multi-dimensional data from all users, analyze data characteristics based on big data analysis algorithms, and allocate corresponding teaching resources to the group. The multi-dimensional data includes basic information, learning ability, and learning preferences. S2. Based on the dynamic grid layout algorithm, the corresponding number of windows are automatically generated on the electronic whiteboard according to the number of groups, and the association between the split-screen windows and the groups is established. S3. Based on the data visualization model, the teaching resources obtained by each group are displayed on the corresponding split-screen window. S4. Detect the user's operation action, determine the split window to which the operation action belongs, and implement the relevant functions of the window according to the operation action; S5. Identify member identities and determine members' level of focus and understanding of the current teaching resources. Based on the difficulty level of the teaching resources, generate a split-screen window adjustment strategy and dynamically adjust the size and position of the split-screen window according to the adjustment strategy.

[0008] Furthermore, in step S1, the collected data is analyzed, and appropriate teaching resources are allocated to the group based on its shared characteristics, including the following steps: Extract shared features from the group; Establish a teaching resource database and provide detailed annotations for each resource; Based on the collaborative filtering algorithm, the cosine similarity between the group and the teaching resources is calculated to perform preliminary matching of teaching resources; The delivery of teaching resources can be dynamically adjusted using Bayesian network models or reinforcement learning algorithms.

[0009] Furthermore, in step S2, after the split-screen window is generated, a unique identifier is assigned to each window, and a mapping table between the window and the group is established. When a user logs into the system and enters the group whiteboard interface, the resources of the corresponding group are displayed in the corresponding split-screen window according to the user's group.

[0010] Furthermore, in step S3, a data visualization model is constructed to display the teaching resources in each window, including the following steps: Utilize front-end technologies to build visual interfaces; Import a visualization library and configure the chart types and related parameters; The teaching resource data is linked to the visualization interface, and when the data changes, it is automatically updated and displayed visually. The visualized teaching resources are integrated into the split-screen area of ​​the whiteboard for display; The integrated and optimized visual teaching resources will be finally presented in the whiteboard split-screen area; The visual interface was optimized based on user feedback.

[0011] Preferably, in step S3, the visually displayed teaching resources include text resources, image resources, video resources, data resources, and interactive resources.

[0012] Furthermore, step S4 includes: Set up an event listener mechanism to detect whether the user has performed any actions; When an operation is detected, the coordinates of the operation location are calculated to determine whether the operation occurred within the split-screen window area; If the operation is determined to be outside the window area, the operation is ignored; if the operation is determined to be within the window area, the corresponding split window is determined based on the operation location, and the relevant functions of the corresponding window are implemented according to the type of operation.

[0013] Furthermore, in step S4, a version control mechanism is introduced to assign a unique version number to each operation on the whiteboard. When a user performs an operation, the operation request will send the current version number to the server. After receiving the request, the server will compare the version number in the request with the latest version number of the resource or region recorded on the server side, and after processing the operation request, synchronize the latest operation result and version number to all relevant clients.

[0014] Furthermore, step S5 includes the following steps: Collect and store facial data from each user, extract facial features based on a convolutional neural network, and compare them with the stored facial features of group members to identify the members' identities; Based on the member identity information obtained by facial recognition, the system uses expression recognition algorithms and eye-tracking technology to determine the member's level of focus and comprehension of the current teaching resources. A teaching resource difficulty assessment model is constructed based on multi-dimensional information of teaching resources. The multi-dimensional information includes the scope of knowledge points covered, question types, and the complexity of problem-solving steps, and the difficulty level is divided into three levels: easy, medium, and hard. A correlation analysis was conducted between the difficulty level of teaching resources and the status of group members to generate strategies for adjusting the size and position of split-screen windows. Based on the layout grid algorithm, the size and position parameters of the split-screen window are recalculated according to the adjustment strategy, and the size and position of the split-screen window are dynamically adjusted according to the parameters.

[0015] Furthermore, assessing a member's level of focus and comprehension of the current teaching resources includes the following steps: Facial expressions are identified using an expression recognition framework. Obtain the geometric features of the eyes, and combine them with camera calibration parameters and a 3D geometric model to calculate the gaze direction of the members; By fusing and analyzing facial expression recognition results and eye-tracking information, and setting different weighting coefficients, the facial expression state and eye gaze location are quantitatively evaluated. Based on the evaluation results, the member's level of focus and understanding of the current teaching resources is judged.

[0016] The beneficial effects of this invention are as follows: This invention provides a method for controlling a grouped whiteboard with automatic split-screen display, relating to the field of split-screen display technology. It dynamically matches and adapts resources to each group by collecting multi-dimensional user information and analyzing data characteristics. Based on a layout grid algorithm, it automatically generates a corresponding number of split-screen windows according to the number of groups and establishes a relationship between windows and groups. Using a data visualization model, resources are displayed in an intuitive and efficient manner in the corresponding windows. Real-time detection of user operations enables simultaneous operation across multiple groups, ensuring real-time information sharing. By identifying member identities and judging members' focus and understanding of the current teaching resources, combined with the difficulty level of the teaching resources, it generates an adjustment strategy for the split-screen windows. Based on this strategy, it dynamically adjusts the size and position of the split-screen windows, significantly improving the smoothness and visualization efficiency of group collaboration, bringing a more efficient and intelligent interactive experience to teaching and teamwork. Attached Figure Description

[0017] Figure 1 This is a flowchart of a grouped whiteboard control method for automatic split-screen display according to the present invention. Detailed Implementation

[0018] Please see Figure 1 As shown, the present invention relates to a method for controlling a grouped whiteboard with automatic split-screen display, comprising the following steps: S1. Collect multi-dimensional data from all users, analyze data characteristics based on big data analysis algorithms, and allocate corresponding teaching resources to the group. The multi-dimensional data includes basic information, learning ability, and learning preferences. Basic Information: When users log in to the system, they are required to fill in basic information such as age, gender, grade, and class, which is then stored in the user information database.

[0019] Learning ability information: By connecting with the school's academic affairs system, data such as users' exam scores, homework completion, and classroom test results can be obtained; at the same time, the system is equipped with entrance tests and periodic ability assessment tests, with test question types covering multiple choice, short answer, and practical questions, to comprehensively evaluate users' mastery of knowledge points in various subjects, problem-solving ability, and learning speed.

[0020] Learning preference information: The VARK learning style questionnaire can be used, which includes questions on visual, auditory, reading and writing, and kinesthetic dimensions. It provides a list of options for subject areas and learning activity formats (such as group discussions, experimental operations, case analysis, etc.). Users can select the content they are interested in. After users complete the questionnaire, the system analyzes their learning preferences based on their answers.

[0021] The collected information is cleaned to remove duplicate and erroneous data, and missing data is processed using methods such as mean imputation and multiple imputation. Non-numerical data is encoded and transformed, such as converting category data such as learning style and interest preference into numerical vectors.

[0022] The collected data is analyzed, and appropriate teaching resources are allocated to the groups based on their shared characteristics. This includes the following steps: Extract shared features from the group; Analyze the multi-dimensional data of all members in the group to extract common features. For example, if all members of group A are deficient in solid geometry in high school mathematics, then when allocating teaching resources to this group, focus on allocating resources such as videos and exercises on solid geometry to improve the group's solid geometry ability.

[0023] Establish a teaching resource database and provide detailed annotations for each resource; Establish a teaching resource library that includes various types of materials such as text, images, audio, video, animation, and interactive exercises. Each resource is labeled in detail, with information including subject, knowledge point, difficulty level, applicable learning style, resource type, and keywords. For example, "High School Mathematics - Functions - Medium Difficulty - Visual, Reading and Writing - Teaching Videos - Function Graphing, Function Properties".

[0024] Based on the collaborative filtering algorithm, the cosine similarity between the group and the teaching resources is calculated to perform preliminary matching of teaching resources; Based on the overall characteristics of the divided groups (such as average learning ability, main learning style, and shared interests), highly relevant teaching resources are selected from the resource library. During matching, a collaborative filtering algorithm is used to calculate the cosine similarity between the group and the resource, prioritizing resources with a similarity greater than 0.7 and a difficulty level that does not fluctuate by more than 10% from the group's average. For example, for groups with strong learning abilities, a predominantly visual learning style, and an interest in geometry, resources such as 3D geometry drawing tutorial videos and dynamic geometric animations are recommended.

[0025] The delivery of teaching resources can be dynamically adjusted using Bayesian network models or reinforcement learning algorithms.

[0026] By setting up tracking points at various stages such as resource browsing and learning task completion, user learning behavior data is collected in real time, including resource browsing time, learning task completion time, answer accuracy rate, and discussion forum content. Using a Bayesian network model, the probability of a group's need for different resources is updated based on real-time data, dynamically adjusting resource recommendations. If a group has a low answer accuracy rate for a certain knowledge point, basic explanation resources and reinforcement exercises for that knowledge point are added; if a group spends a long time browsing a certain type of resource (such as videos) and provides positive feedback, further extension resources of the same type are recommended. Alternatively, a Q-Learning algorithm based on reinforcement learning can be used, where user actions on resources (such as saving or liking) are used as reward signals, and the resource recommendation strategy is updated every 10 minutes. For example, if a group frequently skips a certain video, the recommendation weight of similar videos is reduced.

[0027] All teaching resources matched by a group can be represented in the form of a folder or compressed file to facilitate resource retrieval. The group and the matched teaching resources can be associated and recorded. For example, if group 1 matches teaching resource 1, it can be recorded as "group 1-teaching resource 1".

[0028] S2. Based on the dynamic grid layout algorithm, the corresponding number of windows are automatically generated on the electronic whiteboard according to the number of groups, and the association between the split-screen windows and the groups is established. A dynamic grid layout algorithm is used to automatically calculate the layout of the split-screen windows based on the number of groups, N. When N=1, full-screen display is used; when N=2, the whiteboard is divided into two columns (left and right) or two rows (top and bottom); when N=3, it is divided into three columns or two rows (one row contains two windows); when N≥4, a row-column matrix layout is preferred, such as a 2×2 matrix for N=4 and a 2×3 matrix for N=6. The initial size of each split-screen window is proportionally allocated according to the whiteboard screen size and the number of groups to ensure that the window size is appropriate and the content is clearly visible.

[0029] After the split-screen windows are generated, a unique identifier ID is assigned to each window, and a mapping table between window IDs and group IDs is established. When a user logs into the system and enters the group whiteboard interface, the resources of the user's corresponding group are displayed in the appropriate split-screen window based on the user's group ID. At the same time, the group name or logo is displayed in the window title bar or border for easy user identification. For example, if group 1 is associated with window 1, it is recorded as "group 1-window 1".

[0030] S3. Based on the data visualization model, the teaching resources obtained by each group are displayed on the corresponding split-screen window. To build a data visualization model to display the teaching resources in each window, the following steps are included: Utilize front-end technologies such as HTML5, CSS3, and JavaScript to build visual interfaces; For example, using HTML5 <canvas>element or <svg>Elements are used to draw charts and graphs; CSS3 is used for styling to achieve beautiful interface effects; and JavaScript is used to implement interactive functions and dynamic data updates.

[0031] Import a visualization library and configure the chart types and related parameters; For example, Echarts, D3.js, and Chart.js. Taking Echarts as an example, you only need to prepare data according to its data format requirements, configure the chart type and related parameters, and you can quickly generate various charts. For complex visualization needs, D3.js provides more powerful customization features, which can flexibly manipulate DOM elements according to specific needs to achieve highly customized visualization effects.

[0032] The teaching resource data is linked to the visualization interface, and when the data changes, it is automatically updated and displayed visually. Asynchronous data loading is achieved using AJAX technology, ensuring that the page does not flicker when data is updated. For complex visualizations, such as dynamic chart animations and 3D model displays, WebGL technology is used for rendering. For example, after a student submits their answer, the backend returns the answer result data to the frontend, and the frontend updates the answer accuracy chart in real time through a data binding mechanism.

[0033] The visualized teaching resources are integrated into the split-screen area of ​​the whiteboard for display; To optimize the performance of visualization content and reduce resource loading time and memory usage, for example, images are compressed to reduce image file size; for large data charts, pagination or sampling display is used to improve page rendering speed; and browser caching technology is used to cache loaded resources to avoid duplicate requests.

[0034] The integrated and optimized visual teaching resources will be finally presented in the whiteboard split-screen area; Ensure the interface is aesthetically pleasing, user-friendly, and clearly communicates information. Each group member can view and interact with visual resources within their designated split-screen area, enabling efficient collaborative learning.

[0035] The visual interface was optimized based on user feedback.

[0036] During use, we collect feedback from users (teachers and students) on the visualization effects to understand the problems and needs they encounter when viewing and using the resources. Based on the feedback, we further improve and optimize the visualization scheme and technical implementation to continuously enhance the quality and effectiveness of the visualization of teaching resources.

[0037] Based on the content of the teaching resources matched by the window, and using a data visualization model, different display methods are adopted to present the teaching resources, which include text resources, image resources, video resources, data resources, interactive resources, and other types.

[0038] Text Resources: For shorter texts, a fixed layout is used, highlighting key content with different font colors, sizes, and formatting such as bolding and underlining. For longer texts, scrollable text boxes are used, along with a table of contents for easy navigation. Furthermore, natural language processing technology is employed to extract key information from the text and generate mind maps, graphically representing the text's logical structure. The CodeMirror plugin provides editable text areas and supports Markdown syntax highlighting; for long texts, TreeMap visualization technology is used to display the chapter structure.

[0039] Image Resources: The system automatically adjusts the image scaling ratio based on the image size and whiteboard window size to ensure complete and undistorted image display. For multiple images, a slideshow or thumbnail preview method can be used for display. Users can click on the thumbnail to view a high-resolution version of the image, and image rotation and panning operations are supported.

[0040] Video Resources: An HTML5 video player is embedded in a split-screen window, providing basic control buttons such as play, pause, fast forward, rewind, and volume adjustment, as well as displaying the video playback progress bar and duration. It supports automatic playback and loop playback settings, and also provides video subtitle display and hiding functions.

[0041] Data Resources: Choose the appropriate chart type based on the data type and analysis purpose, such as bar charts for comparing data sizes, line charts for showing data trends, pie charts for illustrating data proportions, and scatter plots for analyzing data correlations. Use visualization libraries such as Echarts and D3.js to generate charts and add data labels, axis labels, legends, and titles to enhance chart readability. Simultaneously, support interactive chart features, such as displaying detailed data information on mouse hover and filtering and drill-down functions by clicking on chart elements.

[0042] Interactive resources: Design a simple and intuitive user interface, rationally arranging operation buttons, input boxes, drop-down menus, and other elements in a split-screen window. For online tests, display information such as question number, remaining time, and number of questions answered in real time; after the user completes the test, immediately display the answer result and score, and provide explanations for incorrect questions and links to review relevant knowledge points.

[0043] During use, the user's learning progress is recorded. For example, if user 1 belongs to group 1, is associated with window 1, is matched with teaching resource 1, and the learning progress is the function part of the teaching resource, then it is recorded as "user 1-group 1-window 1-teaching resource 1-function".

[0044] S4. Detect the user's operation action, determine the split window to which the operation action belongs, and implement the relevant functions of the window according to the operation action; Step S4 includes: Set up an event listener mechanism to detect whether the user has performed any actions; Add a global event listener to the whiteboard page to listen for operation events such as mouse clicks, drags, scroll wheels, and keyboard input, as well as touch events such as touch start, move, and end on touch devices.

[0045] When an operation is detected, the coordinates of the operation location are calculated to determine whether the operation occurred within the split-screen window area; For example, events such as mousedown, mousemove, and mouseup can be bound to an HTML5 Canvas, and the getBoundingClientRect method can be used to determine whether the operation is within the window. For touch devices, OpenCV.js is used to implement gesture recognition, such as two-finger zoom and three-finger window switching. When generating split-screen windows, each window has its corresponding rectangular area, which can be defined by the coordinates of the top-left corner (x, y), width, and height of the window. By comparing the operation position coordinates with the boundary coordinates of the window area, it can be determined whether the operation occurred within that window.

[0046] If the operation is determined to be outside the window area, the operation is ignored; if the operation is determined to be within the window area, the corresponding split window is determined based on the operation location, and the relevant functions of the corresponding window are implemented according to the type of operation.

[0047] In determining whether an operation is within a window, once a window containing the operation's location is found, it can be identified as the split-screen window to which the operation belongs. Typically, each split-screen window is associated with a unique identifier (such as windowId). When traversing windows, if it is determined that an operation is within a certain window, the unique identifier (such as windowId) of that window can be obtained, thus determining the split-screen window to which it belongs.

[0048] User actions include writing annotations, resource operations, window adjustments, and group collaboration.

[0049] Writing and annotation operations: If mouse dragging or touch movement is detected within the split-screen window, and the operation type is writing or annotation, lines are drawn or annotation content is added within the window based on the operation trajectory. The line color, thickness, and style can be customized by the user in the toolbar, and the annotation content supports text input and format editing. Resource operations: When a user clicks on a resource (such as a video play button, image thumbnail, link, etc.) in the split-screen window, the corresponding operation is performed according to the resource type, such as playing a video, viewing a larger image, or jumping to the relevant page. Window adjustment operation: If a user drags the border of the split-screen window, adjust the window size according to the drag direction and distance. When adjusting the window size, ensure that there is a reasonable gap between adjacent windows to avoid overlapping. If a user drags the window title bar, move the window according to the drag distance and target position. When moving the window, restrict the window movement range to within the whiteboard screen. Group collaboration: When users engage in group collaboration operations such as speaking in the discussion area or sharing files within a split-screen window, the system uses the WebSocket protocol to synchronize the operation data to the corresponding split-screen windows of other members in the same group in real time, enabling real-time collaboration within the group.

[0050] The WebSocket protocol is used to enable real-time communication between groups and between groups and the teacher. Each group's actions (such as writing, annotating, and submitting answers) are transmitted to the server in real time via the WebSocket protocol, and then the server pushes them to other relevant clients to ensure that the whiteboard content of all groups and the teacher is updated synchronously.

[0051] In scenarios where multiple users operate simultaneously, a version control mechanism is introduced to prevent data conflicts and ensure orderly execution of operations. The core principle is that each operation generates a version number, and the server processes operation requests sequentially based on these version numbers, guaranteeing consistency and correctness. For example, when group A and group B simultaneously edit the same area, the server processes the requests in chronological order to avoid data overwriting.

[0052] The system assigns a unique version number to each operation on the whiteboard (such as writing, annotation, resource modification, window adjustment, etc.). The version number can be an incrementing numerical sequence, for example, starting from 1 and incrementing by 1 with each operation; or it can be generated using a combination of timestamps and random numbers. In addition to information such as the operation content, operator, and operation time, each operation record also carries a corresponding version number to identify the order and status of the operation.

[0053] When a user performs an operation, the operation request sends the current version number to the server. Upon receiving the request, the server compares the version number in the request with the latest version number of the resource or region recorded on the server. If the version number in the request matches the latest version number on the server, it means that the data has not been modified by other operations when the operation was initiated, and the server will execute the operation, increment the latest version number by 1, and update the version record on the server. If the version number in the request is less than the latest version number on the server, it indicates that other operations have modified the data during the operation's initiation, and the server will reject the operation, returning the latest version information to the client, prompting the user that their operation is based on an outdated data version and that they need to obtain the latest data before proceeding.

[0054] In some situations, multiple operations may not overlap and need to be merged. For example, two users may be writing and annotating in different areas of a whiteboard simultaneously. In this case, the server can intelligently merge these operations based on their location and type.

[0055] After processing the operation request, the server synchronizes the latest operation result and version number to all relevant clients via real-time communication technology (such as WebSocket). Upon receiving the update, the client updates its local whiteboard display and version record to ensure that all users see consistent and up-to-date data. Simultaneously, when a client initiates a new operation request, it will be based on the updated version number to guarantee the correctness and consistency of subsequent operations.

[0056] S5. Identify member identities and determine members' level of focus and understanding of the current teaching resources. Based on the difficulty level of the teaching resources, generate a split-screen window adjustment strategy and dynamically adjust the size and position of the split-screen window according to the adjustment strategy.

[0057] Step S5 includes the following steps: Collect and store facial data from each user, locate the face region in the video frame based on a multi-task cascaded convolutional network, identify the face and extract its features, compare it with the stored facial features of the group members, and identify the member's identity. Specifically, a high-definition camera is integrated into the electronic whiteboard to collect video stream data at a frequency of 25-30 frames per second. The face region in the video frame is quickly located through a multi-task cascaded convolutional network, and all faces in the picture are identified. For the detected faces, a feature extraction model, such as FaceNet, is used to map the face image into a 128-dimensional feature vector. The extracted feature vector is compared with the stored member face feature database. By calculating the cosine similarity or Euclidean distance, when the calculated similarity is greater than the set threshold, the member's identity can be identified.

[0058] Based on the member identity information obtained by facial recognition, the system uses expression recognition algorithms and eye-tracking technology to determine the member's level of focus and comprehension of the current teaching resources. Facial expressions are identified using an expression recognition framework. Key feature points of facial expressions are extracted using facial expression recognition frameworks such as OpenFace or DeepFace. Expression features are quantified by calculating the relative positions and angular changes between these feature points. For example, the angle of the corners of the mouth and the degree of eyebrow furrowing can serve as important criteria for expression recognition. The extracted expression features are then input into a pre-trained expression classification model. The model outputs a probability value for each expression category using a Softmax function, and the category with the highest probability value is taken as the current expression state of the user.

[0059] Obtain the geometric features of the eyes, and combine them with camera calibration parameters and a 3D geometric model to calculate the gaze direction of the members; By detecting key information such as the eye's outline, iris position, and pupil center, the geometric features of the eye are obtained. Based on the eye feature detection results, combined with camera calibration parameters and a 3D geometric model, the member's gaze direction is calculated. By establishing an eyeball model and calculating the relative positional relationship between the pupil center and the corneal reflection point, the direction vector of the gaze in 3D space is deduced. This vector is then intersected with the electronic whiteboard screen plane to determine the coordinates of the member's gaze point on the screen, thereby judging whether the member is paying attention to the split-screen window area where the current teaching resource is located. If the gaze point deviates from the window area for an extended period, it indicates that the member's concentration is low.

[0060] By fusing and analyzing facial expression recognition results and eye-tracking information, and setting different weighting coefficients, the facial expression state and eye gaze location are quantitatively evaluated. Based on the evaluation results, the member's level of focus and understanding of the current teaching resources is judged.

[0061] The focus and comprehension of members on the current teaching resources are quantified into different assessment scores. Each assessment score corresponds to a state. For example, if a member's expression is "confused" and their gaze is off the window for a long time, it is determined that their focus is low and they have difficulty understanding the teaching resources.

[0062] A teaching resource difficulty assessment model is constructed based on multi-dimensional information of teaching resources. The multi-dimensional information includes the scope of knowledge points covered, question types, and the complexity of problem-solving steps, and the difficulty level is divided into three levels: easy, medium, and hard. For text-based resources, natural language processing (NLP) technology is used to analyze the number of technical terms and sentence structure complexity. For exercise-based resources, the difficulty is assessed based on the number of knowledge points involved and the level of logical reasoning required to solve the problem. For experimental operation video resources, the difficulty level is determined based on the number of operation steps and accuracy requirements. The difficulty levels are divided into three categories: easy, medium, and hard, and each resource is assigned a corresponding difficulty tag upon entry into the database. Furthermore, after resources are matched within a group, the system generates a dynamic difficulty adjustment coefficient by considering the group's overall learning ability, the suitability of the resource's difficulty, and the members' focus and comprehension of the teaching resource. This coefficient is used for subsequent window adjustment decisions.

[0063] A correlation analysis was conducted between the difficulty level of teaching resources and the status of group members to generate strategies for adjusting the size and position of split-screen windows. When the resource difficulty is "hard" and most members appear confused, the system determines that the corresponding split-screen window for that group needs to be enlarged to display more supplementary materials or step-by-step explanations. When the resource difficulty is "easy" and members are focused and actively interacting, the window can be appropriately reduced to make room for other groups. Additionally, based on the members' positions in front of the whiteboard, if the group as a whole is close to the left side of the whiteboard, the system can move that group's window to the left for easier member operation.

[0064] Based on the layout grid algorithm, the size and position parameters of the split-screen window are recalculated according to the adjustment strategy, and the size and position of the split-screen window are dynamically adjusted according to the parameters.

[0065] Based on the adjustment strategy, the size and position parameters of the split-screen window are recalculated, and the electronic whiteboard page is rendered and updated in real time using CSS3's transform property or a front-end graphics library (such as D3.js). This achieves animation effects for window scaling and position movement, ensuring a smooth transition and no impact on user experience. After the adjustment is completed, the system continues to monitor the group members' facial information and resource usage in real time, continuously and dynamically optimizing the split-screen window settings to adapt to various changes during the group learning process.

[0066] The above embodiments are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.< / svg> < / canvas>

Claims

1. A method for controlling a grouped whiteboard with automatic split-screen display, characterized in that, Includes the following steps: S1. Collect multi-dimensional data from all users, analyze data characteristics based on big data analysis algorithms, and allocate corresponding teaching resources to the group. The multi-dimensional data includes basic information, learning ability, and learning preferences. S2. Based on the dynamic grid layout algorithm, the corresponding number of windows are automatically generated on the electronic whiteboard according to the number of groups, and the association between the split-screen windows and the groups is established. S3. Based on the data visualization model, the teaching resources obtained by each group are displayed on the corresponding split-screen window. S4. Detect the user's operation action, determine the split window to which the operation action belongs, and implement the relevant functions of the window according to the operation action; S5. Identify member identities and determine members' level of focus and understanding of the current teaching resources. Combine the difficulty level of the teaching resources to generate a split-screen window adjustment strategy. Based on the adjustment strategy, dynamically adjust the size and position of the split-screen window. In step S4, a version control mechanism is introduced to assign a unique version number to each operation on the whiteboard. When a user performs an operation, the operation request will send the current version number to the server. After receiving the request, the server will compare the version number in the request with the latest version number of the resource or region recorded on the server side, and after processing the operation request, synchronize the latest operation result and version number to all relevant clients. Step S5 includes the following steps: Collect and store facial data from each user, extract facial features based on a convolutional neural network, and compare them with the stored facial features of group members to identify the members' identities; Based on the member identity information obtained by facial recognition, the system uses expression recognition algorithms and eye-tracking technology to determine the member's level of focus and comprehension of the current teaching resources. A teaching resource difficulty assessment model is constructed based on multi-dimensional information of teaching resources. The multi-dimensional information includes the scope of knowledge points covered, question types, and the complexity of problem-solving steps, and the difficulty level is divided into three levels: easy, medium, and hard. A correlation analysis was conducted between the difficulty level of teaching resources and the status of group members to generate strategies for adjusting the size and position of split-screen windows. Based on the layout grid algorithm, the size and position parameters of the split-screen window are recalculated according to the adjustment strategy, and the size and position of the split-screen window are dynamically adjusted according to the parameters.

2. The automatic split-screen display grouped whiteboard control method according to claim 1, characterized in that, In step S1, the collected data is analyzed, and appropriate teaching resources are allocated to the groups based on their shared characteristics. This includes the following steps: Extract shared features from the group; Establish a teaching resource database and provide detailed annotations for each resource; Based on the collaborative filtering algorithm, the cosine similarity between the group and the teaching resources is calculated to perform preliminary matching of teaching resources; The delivery of teaching resources can be dynamically adjusted using Bayesian network models or reinforcement learning algorithms.

3. The automatic split-screen display grouped whiteboard control method according to claim 1, characterized in that, In step S2, after the split-screen window is generated, a unique identifier is assigned to each window, and a mapping table between the window and the group is established. When a user logs into the system and enters the group whiteboard interface, the resources of the corresponding group are displayed in the corresponding split-screen window according to the user's group.

4. The method for controlling a grouped whiteboard with automatic split-screen display according to claim 1, characterized in that, In step S3, a data visualization model is constructed to display the teaching resources in each window, including the following steps: Utilize front-end technologies to build visual interfaces; Import a visualization library and configure the chart types and related parameters; The teaching resource data is linked to the visualization interface, and when the data changes, it is automatically updated and displayed visually. The visualized teaching resources are integrated into the split-screen area of ​​the whiteboard for display; The integrated and optimized visual teaching resources will be finally presented in the whiteboard split-screen area; The visual interface was optimized based on user feedback.

5. The automatic split-screen display grouped whiteboard control method according to claim 4, characterized in that, In step S3, the visually displayed teaching resources include text resources, image resources, video resources, data resources, and interactive resources.

6. The method for controlling a grouped whiteboard with automatic split-screen display according to claim 1, characterized in that, Step S4 includes: Set up an event listener mechanism to detect whether the user has performed any actions; When an operation is detected, the coordinates of the operation location are calculated to determine whether the operation occurred within the split-screen window area; If the operation is determined to be outside the window area, the operation is ignored; if the operation is determined to be within the window area, the corresponding split window is determined based on the operation location, and the relevant functions of the corresponding window are implemented according to the type of operation.

7. The method for controlling a grouped whiteboard with automatic split-screen display according to claim 1, characterized in that, Assessing a member's level of focus and comprehension of the current teaching resources includes the following steps: Facial expressions are identified using an expression recognition framework. Obtain the geometric features of the eyes, and combine them with camera calibration parameters and a 3D geometric model to calculate the gaze direction of the members; By fusing and analyzing facial expression recognition results and eye-tracking information, and setting different weighting coefficients, the facial expression state and eye gaze location are quantitatively evaluated. Based on the evaluation results, the member's level of focus and comprehension of the current teaching resources is determined.

Images