Immersive ideological and political education system and method based on VR and red building digital twinning

Abstract

The invention relates to the field of virtual reality technology, digital twinning technology and ideological and political education, in particular to an immersive ideological and political education system and method based on VR and red building digital twinning. VR, deep learning and digital twinning technologies are fused, and an immersive learning environment is created; the system comprises VR simulation which presents a full-digital red building and supports space interaction; a deep learning algorithm is adopted to accurately restore the appearance, environment and internal details of the building; the ideological and political knowledge point library is connected with the VR and modeling system and provides rich teaching content; an ideological and political feedback mechanism for collecting user emotions and dynamically adjusting teaching; the system not only realizes high-precision digitalization of the red building, but also enhances the participation sense of learners through immersive experience, innovates an ideological and political education mode, and improves the teaching effect.

Description

Technical Field

[0001] The present invention relates to the fields of virtual reality technology, digital twin technology, and ideological and political education, and particularly relates to an immersive ideological and political education system and method based on VR and digital twin of red buildings. Background Art

[0002] With the development of information technology, the application of virtual reality (VR) technology and digital twin technology in the field of education is becoming increasingly widespread. Traditional ideological and political education mainly relies on methods such as classroom lectures and text reading, and there are problems such as insufficient interactivity, lack of immersion, and difficulty in quantifying teaching effects. Introducing VR technology into ideological and political education can create an immersive learning experience and enhance teaching interactivity and attractiveness.

[0003] The existing VR education systems mainly have the following deficiencies: First, the accuracy of digital models is not high, and it is difficult to restore the historical details of red buildings; second, there is a lack of professional design for ideological and political education content; third, the interaction methods are single, and it is difficult to meet the in-depth needs of ideological and political education; fourth, there is a lack of real-time collection and analysis of users' emotional feedback, and it is difficult to adjust teaching strategies according to the learners' states.

[0004] Currently, there is no professional system at home and abroad for the deep integration of digital twin of red buildings and ideological and political education. Therefore, there is an urgent need to develop an immersive ideological and political education system based on VR and digital twin of red buildings, and through high-precision digital modeling, in-depth interactive experience, and emotional feedback analysis, to enhance the immersion and teaching effect of ideological and political education. Summary of the Invention

[0005] The purpose of the present invention is to provide an immersive ideological and political education system and method based on VR and digital twin of red buildings, aiming to solve the problems of insufficient interactivity, lack of immersion, and difficulty in quantifying teaching effects in traditional ideological and political education.

[0006] The present invention proposes an immersive ideological and political education system based on VR and digital twin of red buildings, including:

[0007] A VR simulation system for providing a fully digital model of a red building to a user and enabling the user to interact with the model spatially;

[0008] A deep learning algorithm modeling system connected to the VR simulation system for extracting building appearance, environment, internal exhibits, and scene elements and performing digital simulation to build a fully digital virtual environment of a red building;

[0009] An ideological and political knowledge point system connected to the VR simulation system and the deep learning algorithm modeling system for storing ideological and political knowledge points;

[0010] The ideological and political feedback system is connected to the VR simulation system and is used to collect emotional data based on user behavior, establish an emotional model, and dynamically adjust the teaching method.

[0011] Preferably, the VR simulation system includes:

[0012] A scene construction module for realizing scene construction and scene interaction; wherein, the virtual scene is a twin digital environment designed at a 1:1 ratio according to the real scene of the red building relying on digital technologies such as virtual reality technology and modeling technology;

[0013] A scene interaction module, connected to the scene construction module, for monitoring the user's body position, footsteps and head movements, controlling the virtual camera view; responding to the user's trigger of the switch device, controlling the lighting, sound effects and scene transformation; and identifying the user's trigger of the interaction device, calling the associated knowledge points and performing scene linkage.

[0014] Preferably, the deep learning algorithm modeling system includes:

[0015] An architectural appearance modeling module for inputting building pictures into a convolutional neural network, calculating the pixel edge intensity through edge detection, extracting pixels and calculating histogram features, and constructing a digital model of the architectural appearance;

[0016] A scene element modeling module for extracting scene elements from historical pictures, using transfer learning methods to construct a scene element library, reasoning about historical elements, and establishing a full digital model of scene elements;

[0017] An internal scene construction module for using transfer learning methods to extract the internal scene elements of buildings in different historical eras and construct a full digital model of the internal scene.

[0018] Preferably, the ideological and political knowledge point system includes:

[0019] A knowledge point acquisition module for obtaining ideological and political knowledge points through manual collation, classifying the knowledge points and storing them;

[0020] A push plan design module for classifying the full digital building model according to the actual scene during the virtual scene construction and corresponding it to the ideological and political knowledge points one by one;

[0021] A knowledge point application module for matching with knowledge points according to the user's interaction actions during scene interaction and conducting ideological and political education through push messages.

[0022] Preferably, the ideological and political feedback system includes:

[0023] An emotion collection module, which is used to regularly capture the user's facial images through a camera and collect voice data through a microphone, and process the data to obtain emotion data;

[0024] An emotion model building module, connected to the emotion collection module, which is used to use a deep neural network, take emotion data as a training set, and build an emotion model through iterative training;

[0025] A feedback mechanism design module, connected to the emotion model building module, which is used to predict and give feedback on the user's actions based on the emotion model, and design corresponding teaching feedback for the user according to the analysis results.

[0026] Preferably, the classifier used by the emotion model building module includes at least one of KNN, SVM, RandomForest, CNN, RNN, and LSTM.

[0027] Preferably, the feedback mechanism design module includes:

[0028] An emotion recognition unit, which is used to recognize the user's expression state. When the user's expression is frowning or shaking the head, it is judged as a confused state. When the user's expression is smiling or nodding, it is judged as an approval state;

[0029] A teaching adjustment unit, connected to the emotion recognition unit, which is used to pause the current content playback and push explanatory content when a confused state is detected, and continue playing and provide in-depth knowledge points when an approval state is detected;

[0030] An interactive scenario generation unit, connected to the teaching adjustment unit, which is used to create personalized random interactive scenarios based on user feedback to improve teaching participation.

[0031] Preferably, the system further includes:

[0032] A red scenario database module, connected to the VR simulation system, the ideological and political knowledge point system, and the ideological and political feedback system, which is used to store red building digital scenarios, ideological and political knowledge point content, and teaching materials for different emotional states;

[0033] A teaching evaluation module, connected to the ideological and political feedback system, which is used to record the user's learning behaviors and reactions, analyze the degree of knowledge mastery, and generate a teaching evaluation report.

[0034] Preferably, the system includes a front-end teaching terminal and a back-end control terminal;

[0035] The front-end teaching terminal includes VR glasses, an infrared sensor, and a VR handle;

[0036] The back-end control terminal includes a memory, a processing unit, an input device, an output device, and a network communication unit;

[0037] The VR glasses include a left lens, a right lens, a processor, a gyroscope, an electromagnetic touch controller, a microphone, a USB interface, and an audio playback device;

[0038] The infrared sensor is installed on the VR glasses and the VR handle for sensing the device position and steering.

[0039] An immersive ideological and political education method based on VR and digital twin of red buildings includes:

[0040] Collect red building pictures, use the building pictures at different angles as training samples, learn the building appearance using a convolutional neural network, and build a fully digital model;

[0041] Extract key elements from historical pictures, use the algorithm of transfer learning to build a scene element library including elements such as text, flags, slogans, people, items, etc., classify the elements, match the elements with the fully digital model, restore the building scene, and build a digital twin scene of the red building;

[0042] In the built scene, push animation introductions to push ideological and political knowledge points, and collect user emotion data through a camera and a sound collection device;

[0043] Learn the emotion data through a deep neural network, establish an emotion model, and adjust the teaching method according to user feedback.

[0044] The present invention has the following beneficial effects:

[0045] 1. By combining VR technology with digital twin technology of red buildings, create a highly immersive ideological and political education environment, enhancing the experience and participation of learners;

[0046] 2. Adopt deep learning algorithms for modeling the building appearance, environment, and scene elements to achieve high-precision 1:1 digital restoration of red buildings;

[0047] 3. Establish an accurate mapping between ideological and political knowledge points and digital scenes to achieve contextual and scenario-based knowledge transfer;

[0048] 4. By collecting emotions and constructing an emotion model, grasp the emotional state of learners in real time and dynamically adjust teaching strategies;

[0049] 5. Form a teaching closed-loop system of "perception - analysis - adjustment - feedback", greatly improving the pertinence and effectiveness of ideological and political education. Description of the Drawings

[0050] Figure 1 It is the overall architecture diagram of the immersive ideological and political education system based on VR and digital twin of red buildings of the present invention;

[0051] Figure 2 This is a schematic structural diagram of the VR simulation system of the present invention;

[0052] Figure 3 This is a schematic structural diagram of the deep learning algorithm modeling system of the present invention;

[0053] Figure 4 This is a schematic structural diagram of the ideological and political knowledge point system of the present invention;

[0054] Figure 5 This is a schematic structural diagram of the ideological and political feedback system of the present invention;

[0055] Figure 6 This is a flowchart of the immersive ideological and political education method based on VR and digital twin of red buildings of the present invention. Specific implementation manners

[0056] Please refer to the appendix Figure 1-6 , the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to 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. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0057] Embodiment 1: Immersive ideological and political education system based on VR and digital twin of red buildings

[0058] As Figure 1 shown, the immersive ideological and political education system based on VR and digital twin of red buildings includes a VR simulation system 1, a deep learning algorithm modeling system 2, an ideological and political knowledge point system 3, and an ideological and political feedback system 4.

[0059] The VR simulation system 1 is used to provide a fully digital model of a red building to the user, enabling the user to interact with the model spatially. The deep learning algorithm modeling system 2 is connected to the VR simulation system 1 and is used to extract building appearance, environment, internal exhibits, and scene elements and perform digital simulation to build a fully digital virtual environment of the red building. The ideological and political knowledge point system 3 is connected to the VR simulation system 1 and the deep learning algorithm modeling system 2 and is used to store ideological and political knowledge points. The ideological and political feedback system 4 is connected to the VR simulation system 1 and is used to collect emotion data according to user behavior, establish an emotion model, and dynamically adjust the teaching method.

[0060] As Figure 2As shown in the figure, the VR simulation system 1 includes a scene construction module 11 and a scene interaction module 12. The scene construction module 11 is used to implement scene construction and scene interaction. Among them, the virtual scene is a twin digital environment designed 1:1 according to the real scene by relying on digital technologies such as virtual reality technology and modeling technology. Preferably, the scene construction module 11 of the present invention also includes the establishment of digital models such as scene elements and interaction devices, uses a computer to digitally restore the real scene, and uses digital technology for design. The scene interaction module 12 is connected to the scene construction module 11, and is used to monitor the user's body position, footsteps and head movements, control the virtual camera view; respond to the user triggering the switch device, control the lighting, sound effects and scene transformation; and identify the user triggering the interaction device, call the associated knowledge points and perform scene linkage.

[0061] Specifically, the scene interaction module 12 uses an action recognition algorithm to judge the user's interaction behavior. This algorithm is based on infrared sensor and gyroscope data, and identifies user actions through feature extraction and pattern matching. The action recognition algorithm can be expressed as:

[0062]

[0063] Among them, A(m) is the recognized action type, m is the action data, F(m) is the extracted feature vector, and P(a i |F(m)) is the posterior probability of action a i under the condition of a given feature vector. When the recognition probability exceeds a preset threshold (generally set to 0.85, an empirical value determined based on a large amount of experimental data), the system confirms that the user has performed a specific action and triggers the corresponding interaction response.

[0064] As Figure 3 shown in the figure, the deep learning algorithm modeling system 2 includes an architectural appearance modeling module 21, a scene element modeling module 22, and an internal scene construction module 23. The architectural appearance modeling module 21 is used to input building pictures into a convolutional neural network, calculate the pixel edge intensity through edge detection, extract pixels and calculate histogram features, and construct a digital model of the architectural appearance. The scene element modeling module 22 is used to extract scene elements from historical pictures, use the transfer learning method to construct a scene element library, reason about historical elements, and establish a fully digital model of scene elements. The internal scene construction module 23 is used to use the transfer learning method to extract the internal scene elements of buildings in different historical eras and construct a fully digital model of the internal scene.

[0065] Preferably, the architectural appearance modeling module 21 uses a convolutional neural network for architectural appearance feature extraction and modeling. This convolutional neural network includes multiple convolutional layers, pooling layers and fully connected layers, and is used to extract features from building pictures. The edge detection algorithm uses the Sobel operator to calculate the pixel edge intensity:

[0066]

[0067] Among them, I is the input image, and G x and G y are the gradients in the horizontal and vertical directions respectively, and G is the edge intensity. Using the edge intensity as the weight, pixels are extracted to calculate the histogram features, and a feature model is constructed.

[0068] The scene element modeling module 22 uses transfer learning methods to extract features from a pre-trained model. The transfer learning model is based on pre-trained networks such as ResNet-50 or VGG-16 and is fine-tuned to adapt to specific scene element recognition tasks. The feature extraction is expressed as:

[0069] F e = φ(I e ; θ),

[0070] where F e is the scene element feature vector, I e is the scene element image, φ is the feature extraction function, and θ is the pre-trained model parameter.

[0071] As Figure 4 shown, the ideological and political knowledge point system 3 includes a knowledge point acquisition module 31, a push plan design module 32, and a knowledge point application module 33. The knowledge point acquisition module 31 is used to obtain ideological and political knowledge points through manual collation, classify the knowledge points, and store them. The push plan design module 32 is used to classify the fully digital building model according to the actual scene during the virtual scene construction, and correspond to the ideological and political knowledge points one by one. The knowledge point application module 33 is used to match with the knowledge points according to the user interaction actions during the scene interaction, and conduct ideological and political education through push messages.

[0072] Preferably, the knowledge point acquisition module 31 classifies the ideological and political knowledge points into three major categories: location category, time category, and main event category. The location category includes the names of the exhibition halls where the statues of historical greats are located, the locations of relevant revolutionary meetings, etc.; the time category includes important time nodes such as the meeting time, the birthdays of relevant greats, etc.; the main event category includes core historical events such as the meeting theme, the revolutionary theme, etc. The knowledge point classification adopts a hierarchical index structure for easy quick retrieval and association.

[0073] The push plan design module 32 adopts a content recommendation algorithm based on location and interest. This algorithm comprehensively considers the user location information and the interest model to calculate the knowledge point push priority:

[0074]

[0075] where P(k i ) is the knowledge point k iPush priority, The user location u Location of knowledge points The similarity, S i nt(I u ,k i ) is the user interest model I u With knowledge point k i The matching degree is α, which is a weight parameter and is generally taken as 0.7, an empirical value determined based on user interaction data.

[0076] like Figure 5 As shown, the ideological and political feedback system 4 includes an emotion collection module 41, an emotion model building module 42 and a feedback mechanism design module 43. The emotion collection module 41 is used to capture the user's facial image through the camera and collect voice data through the microphone at regular intervals, and process the data to obtain emotion data. The emotion model building module 42 is connected to the emotion collection module 41, and is used to use a deep neural network, with emotion data as a training set, to build an emotion model through iterative training. The feedback mechanism design module 43 is connected to the emotion model building module 42, and is used to predict and feedback user actions based on the emotion model, and design corresponding teaching feedback for the user according to the analysis results.

[0077] Preferably, the emotion acquisition module 41 uses a facial expression recognition algorithm and a speech emotion analysis algorithm. Facial expression recognition extracts facial features based on a convolutional neural network, and classifies and recognizes seven basic expressions (happy, sad, angry, fearful, disgusted, surprised, and neutral) through a multi-layer perceptron. Speech emotion analysis uses MFCC (Mel-frequency cepstral coefficients) to extract speech features, and combines LSTM networks for temporal emotion modeling.

[0078] The classifier used by the emotion model building module 42 includes at least one of KNN, SVM, Random Forest, C NN, RNN and LSTM. In this embodiment, an integrated learning method is mainly used to integrate the results of multiple classifiers to improve the accuracy of emotion recognition. The integrated learning model is expressed as:

[0079]

[0080] Among them, E(x) is the output of the integrated model, C i (x) is the output of the i-th base classifier, w i is the corresponding weight, satisfying The weight of each classifier is dynamically adjusted according to the performance on the validation set, and the better the performance, the greater the weight.

[0081] The feedback mechanism design module 43 includes an emotion recognition unit 431, a teaching adjustment unit 432, and an interactive scenario generation unit 433. The emotion recognition unit 431 is used to recognize the user's facial expression state. When the user's facial expression is frowning or shaking the head, it is judged as a confused state. When the user's facial expression is smiling or nodding, it is judged as an approval state. The teaching adjustment unit 432 is connected to the emotion recognition unit 431 and is used to pause the current content playback and push explanatory content when a confused state is detected, and continue playback and provide in-depth knowledge points when an approval state is detected. The interactive scenario generation unit 433 is connected to the teaching adjustment unit 432 and is used to create personalized random interactive scenarios based on user feedback to improve teaching participation.

[0082] In addition, the system of the present invention further includes a red scenario database module 5 and a teaching evaluation module 6. The red scenario database module 5 is connected to the VR simulation system 1, the ideological and political knowledge point system 3, and the ideological and political feedback system 4, and is used to store digital red building scenes, ideological and political knowledge point content, and teaching materials for different emotional states. The teaching evaluation module 6 is connected to the ideological and political feedback system 4 and is used to record the user's learning behaviors and reactions, analyze the degree of knowledge mastery, and generate a teaching evaluation report.

[0083] Preferably, the system of the present invention includes a front-end teaching terminal and a back-end control terminal. The front-end teaching terminal includes a VR headset, an infrared sensor, and a VR controller. The back-end control terminal includes a memory, a processing unit, an input device, an output device, and a network communication unit. The VR headset includes a left lens, a right lens, a processor, a gyroscope, an electromagnetic touch controller, a microphone, a USB interface, and an audio playback device. The infrared sensors are installed on the VR headset and the VR controller and are used to sense the device position and turning.

[0084] Embodiment 2: An immersive ideological and political education method based on VR and digital twin of red buildings

[0085] As Figure 6 shown, an immersive ideological and political education method based on VR and digital twin of red buildings includes the following steps:

[0086] Step S1: Collect red building pictures, use the building pictures from different angles as training samples, and use a convolutional neural network to learn the building appearance and build a fully digital model;

[0087] Preferably, the specific implementation process of using a convolutional neural network to learn the building appearance and construct an appearance model in this step is:

[0088] (1) Put the red building pictures into the convolutional neural network, perform edge detection on the pictures, and calculate the edge intensity of the pixels;

[0089] (2) Take the edge intensity as the weight, extract pixels, calculate the pixel histogram features, and establish a feature model;

[0090] (3) Use the histogram matching algorithm to find the training pictures with the same histogram and classify the building appearance;

[0091] (4) Judge whether there is historical building appearance information. If it exists, go to step (6); if not, go to step (5);

[0092] (5) Use the image segmentation algorithm to segment the building appearance, obtain the pixels similar to the feature model, add the similar pictures to the convolutional neural network for extended training, and mark the categories;

[0093] (6) Perform matching according to the feature model and build a fully digital building model.

[0094] Step S2: Extract key elements from historical pictures, use the algorithm of transfer learning to build a scene element library, including elements such as text, flags, slogans, people, items, etc., classify the elements, use the fully digital model to match with the scene elements, restore the building scene, and build a digital twin scene of the red building;

[0095] Preferably, the specific operation of using the transfer learning method to build a fully digital scene model in this step is as follows:

[0096] (1) Build a scene image library, select the elements of the scene from historical pictures, and classify them by category;

[0097] (2) Use the pre-trained model to extract the picture features and build a scene element library;

[0098] (3) Match the scene elements with the building appearance, obtain the matching parameters, and build a fully digital scene model.

[0099] In the present invention, the transfer learning adopts a fine-tuning strategy, that is, on the basis of the pre-trained deep network, the parameters of the first few layers are kept unchanged, and only the parameters of the last few layers are fine-tuned to adapt to the new scene element recognition task. The transfer learning process can be expressed as:

[0100]

[0101] Among them, θ is the model parameter, L is the loss function, f(x i ; θ) is the prediction of the model for the input x i , y i is the true label, R(θ) is the regularization term, λ is the regularization coefficient, generally taking values between 0.001 and 0.01, and is determined according to the data scale and model complexity.

[0102] Step S3: In the constructed scenario, introduce ideological and political knowledge points through push animations, and collect users' emotional data through cameras and sound collection devices;

[0103] Preferably, the specific operation of carrying out ideological and political education in this step is as follows:

[0104] (1) Match the fully digital building model with the scene element model, and build a fully digital scene according to the increase or decrease of scene elements to complete the environmental construction;

[0105] (2) Classify ideological and political knowledge points according to the environmental construction and match them one by one;

[0106] (3) When the user interacts with the fully digital scene, call ideological and political knowledge points according to the interaction actions;

[0107] (4) Push animations to the user to teach the knowledge points and display them in the scene;

[0108] (5) Collect users' emotional data, including facial images and audio data.

[0109] Step S4: Use a deep neural network to learn the emotional data, establish an emotion model, and adjust the teaching method according to the user's feedback.

[0110] Preferably, the specific execution process of adjusting the teaching method in this step is as follows:

[0111] (1) Extract the features of the facial images and audio data, recognize the emotions, and build an emotion dataset;

[0112] (2) Use a deep neural network to extract the hidden knowledge of the emotional data and establish an emotion model;

[0113] (3) Predict the user's feedback based on the facial expressions and audio, and optimize the teaching.

[0114] In the process of emotion analysis, the present invention adopts a multi-modal fusion strategy, combines facial expression and speech emotion features, and improves the emotion recognition accuracy. The multi-modal fusion model is expressed as:

[0115] F fusion =G([F face ,F voice ),

[0116] where F fusion is the fusion feature, F face is the facial expression feature, F voice is the speech emotion feature, G is the fusion function, and the feature fusion can be realized by using an attention mechanism or a gated unit.

[0117] When the system recognizes that the user is in a confused state (such as the frowning index exceeding 0.75 or the head-shaking frequency exceeding 3 times every 5 seconds), it will pause the current content playback and push more detailed explanatory content. When it recognizes that the user is in an agreeing state (such as the smiling index exceeding 0.8 or the nodding frequency exceeding 2 times every 10 seconds), the system will continue to play and push more in-depth knowledge point content. These threshold parameters are obtained through a large number of user experiments and can accurately reflect the user's emotional state.

[0118] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention accordingly. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of the present invention, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present invention.

Claims

1. The immersive ideological and political education system based on VR and red building digital twins is characterized by: include: VR simulation system, used to provide users with a fully digital model of the red building, allowing users to interact with the model in space; A deep learning algorithm modeling system is connected to the VR simulation system to extract the building appearance, environment, internal exhibits and scene elements and perform digital simulation to build a fully digital red building virtual environment; An ideological and political knowledge point system, connected to the VR simulation system and the deep learning algorithm modeling system, for storing ideological and political knowledge points; The ideological and political feedback system is connected to the VR simulation system and is used to collect emotional data based on user behavior, establish an emotional model, and dynamically adjust the teaching method.

2. The immersive ideological and political education system based on VR and red building digital twins according to claim 1 is characterized in that: The VR simulation system comprises: The scene building module is used to realize scene building and scene interaction. The virtual scene is a twin digital environment designed by 1:1 red buildings according to real scenes based on digital technologies such as virtual reality technology and modeling technology. The scene interaction module is connected to the scene building module and is used to monitor the user's posture, footsteps and head movements, control the virtual camera's viewing angle; respond to the user triggering the switch device, control the lighting, sound effects and scene changes; and recognize the user triggering the interactive device, call the associated knowledge points and perform scene linkage.

3. The immersive ideological and political education system based on VR and red building digital twins according to claim 1 is characterized in that: The deep learning algorithm modeling system includes: The building appearance modeling module is used to input the building image into the convolutional neural network, calculate the pixel edge strength through edge detection, extract pixels and calculate the histogram features, and build a digital model of the building appearance; The scene element modeling module is used to extract scene elements from historical pictures, build a scene element library using transfer learning methods, infer historical elements, and establish a fully digital model of scene elements; The internal scene construction module is used to extract the internal scene elements of buildings from different historical eras using the transfer learning method to construct a fully digital model of the internal scene.

4. The immersive ideological and political education system based on VR and red building digital twins according to claim 1 is characterized in that: The ideological and political knowledge point system includes: The knowledge point acquisition module is used to obtain ideological and political knowledge points through manual sorting, and store them after classification; Push solution design module, which is used to classify the fully digital building models according to the actual scenes when building virtual scenes, and correspond them one by one with ideological and political knowledge points; The knowledge point application module is used to match knowledge points according to user interaction actions in scene interactions, and conduct ideological and political education by pushing messages.

5. The immersive ideological and political education system based on VR and red building digital twins according to claim 1 is characterized in that: The ideological and political feedback system includes: The emotion collection module is used to capture the user's facial image through the camera and collect voice data through the microphone at regular intervals, and process the data to obtain emotion data; An emotion model building module, connected to the emotion acquisition module, is used to build an emotion model by using a deep neural network and emotion data as a training set through iterative training; The feedback mechanism design module is connected to the emotion model building module and is used to predict and provide feedback on user actions based on the emotion model, and to design corresponding teaching feedback for the user according to the analysis results.

6. The immersive ideological and political education system based on VR and red building digital twins according to claim 5 is characterized in that: The classifier used in the emotion model building module includes at least one of KNN, SVM, RandomForest, CNN, RNN and LSTM.

7. The immersive ideological and political education system based on VR and red building digital twins according to claim 5 is characterized in that: The feedback mechanism design module includes: Emotion recognition unit, used to recognize the user's facial expression state. When the user's expression is frowning or shaking his head, it is judged as a confused state. When the user's expression is smiling or nodding, it is judged as an approval state. A teaching adjustment unit, connected to the emotion recognition unit, for pausing the current content playback and pushing explanatory content when a confused state is detected, and continuing to play and provide deepening knowledge points when an identification state is detected; The interactive plot generation unit is connected to the teaching adjustment unit and is used to create personalized random interactive plots based on user feedback to improve teaching participation.

8. The immersive ideological and political education system based on VR and red building digital twins according to claim 1 is characterized in that: The system further comprises: A red scene database module is connected to the VR simulation system, the ideological and political knowledge point system and the ideological and political feedback system, and is used to store red building digital scenes, ideological and political knowledge point content and teaching materials for different emotional states; The teaching evaluation module is connected to the ideological and political feedback system and is used to record user learning behaviors and reactions, analyze knowledge mastery, and generate teaching evaluation reports.

9. The immersive ideological and political education system based on VR and red building digital twins according to claim 1 is characterized in that: The system includes a front-end teaching terminal and a back-end control terminal; The front-end teaching terminal includes VR glasses, infrared sensors and VR handles; The back-end control terminal includes a memory, a processing unit, an input device, an output device and a network communication unit; The VR glasses include a left eye lens, a right eye lens, a processor, a gyroscope, an electromagnetic touch controller, a microphone, a USB interface and an audio playback device; The infrared sensor is installed on the VR glasses and the VR handle to sense the position and direction of the device.

10. An immersive ideological and political education method based on VR and red building digital twins, using the system described in any one of claims 1 to 9, characterized in that: include: Collect pictures of red buildings, use pictures of buildings from different angles as training samples, use convolutional neural networks to learn the appearance of buildings, and build a fully digital model; Extract key elements from historical pictures and use transfer learning algorithms to build a scene element library, including text, flags, slogans, people, objects and other elements. Classify the elements and use fully digital models to match the scene elements to restore the building scene and build a digital twin scene of red buildings. In the constructed scenario, ideological and political knowledge points are pushed through push animation introductions, and user emotional data is collected through cameras and sound collection devices; Through deep neural networks, emotional data is learned, emotional models are established, and teaching methods are adjusted according to user feedback.

Images