A Virtual Reality-based Interactive Experience Method for Automobile Exhibitions

Abstract

The present invention belongs to the application of virtual reality technology in the field of automobile exhibitions, and specifically relates to a virtual reality-based interactive experience method for automobile exhibitions. First, a virtual exhibition hall and three-dimensional models of vehicle models are constructed and optimized, enabling users to view the vehicle exterior and internal structure immersively; then, browsing behaviors are recorded, and personalized vehicle model recommendations are generated through analysis of preferences by a recommendation algorithm; after that, after the user selects a vehicle, configuration selection and real-time rendering display can be carried out, and test drive can also be selected. The system loads different driving scenarios to simulate the control experience; finally, a gaze tracking technology based on a clustering algorithm captures the user's attention area to generate a visual heat map to identify key preferences, providing data support for marketing and exhibition planning. This method improves the realism of virtual display, enhances interactivity, enriches the test drive experience, and accurately captures user preferences through multiple technologies, effectively improving the deficiencies of the existing automobile exhibition display methods.

Description

Technical Field

[0001] The present invention belongs to the application of virtual reality technology in the field of automobile exhibitions, and particularly relates to a method for interactive experience of automobile exhibitions based on virtual reality. Background Art

[0002] With the continuous development of virtual reality technology, immersive interactive experiences have gradually been introduced into automobile exhibitions. However, there are many problems with the current existing display methods for automobile exhibitions, making it difficult to meet the growing demand for in-depth experiences from users. There is a lack of immersion, unable to let users feel the charm of automobiles as if they were on the scene; the personalized recommendation function is missing, and it cannot accurately provide model recommendations according to users' preferences; the interactivity is limited, and the interaction between users and the display content is not rich and diverse enough. These problems limit the ability of automobile exhibitions in attracting users, enhancing user experience, and providing effective data support for subsequent marketing and exhibition planning. The present invention aims to solve these problems. By combining technologies such as 3D modeling, personalized recommendation, and virtual test drive simulation, it provides a more immersive experience of vehicle display and configuration. At the same time, it uses eye-tracking technology to analyze user preferences, providing new ideas and methods for the development of automobile exhibitions. Summary of the Invention

[0003] The present invention proposes a method for interactive experience of automobile exhibitions based on virtual reality in view of the existing technical problems.

[0004] To achieve the above object, the technical solution adopted by the present invention is as follows, including the following steps:

[0005] S1. First, use 3D modeling technology to construct a virtual exhibition hall and 3D models of the exhibited vehicles in the hall, and optimize the models to enable users to immerse themselves in viewing the appearance and internal structure of the exhibited vehicles;

[0006] S2. Secondly, record the user's browsing behavior, analyze the preferences through an improved recommendation algorithm, and generate personalized vehicle recommendations in real time for display within the user's field of view for further viewing;

[0007] S3. Then, after the user selects a vehicle model, the system provides color, wheel hub, and seat configuration options and renders and displays them in real time. The personalized configuration can be saved and used for sales communication;

[0008] S4. When the user selects a test drive, the system loads different driving scenarios, simulates the control experience, and feedbacks the acceleration and steering effects, enabling the user to feel the vehicle performance;

[0009] S5. Finally, use eye-tracking technology based on clustering algorithms to capture the specific areas that the user is interested in and generate a visual heat map to identify the key points of user preferences, providing data support for subsequent marketing and exhibition planning;

[0010] In the step S1, the specific implementation of constructing the virtual exhibition hall and the 3D model of the exhibited vehicle types in the exhibition hall by using 3D modeling technology and optimizing the model is as follows:

[0011] S11. First, use the refined modeling technology to perform 3D modeling on the exhibition hall environment and the vehicle. The model is divided into an external geometric model and an internal structure model. The overall model M is expressed as: M = M ext + M int , where M ext represents the vehicle external model, and M int represents the vehicle internal model;

[0012] S12. Then, apply the detail enhancement technology to the vehicle model M. Increase the polygon details through the surface subdivision method to make the surface smoother;

[0013] S13. Then, use high-resolution texture maps to perform texture mapping on the model surface. The texture maps are expressed as T ext and T int , corresponding to the external and internal structures respectively. After mapping, the texture of the model M is: M = M ext · T ext + M int · T int And use the physically based lighting model to enhance the realism. The lighting equation is: L(o) = k d · (N · L i ) · I i + k s · (R · V) n · I i , where L(o) is the output brightness, k d , k s are the diffuse and specular reflection coefficients respectively, N is the normal vector, L i is the incident light, R is the reflection vector, V is the viewing vector, I i is the incident light intensity, and n is the specular coefficient;

[0014] S14. In addition, implement real-time rendering based on the deferred shading technology. Render the geometric information, normal vector, and texture data to the buffer first, and then perform the lighting calculation to optimize the rendering efficiency. The target of the real-time rendering frame rate is F ≥ 90 FPS to meet the fluency of the immersive experience;

[0015] S15. Finally, use the spatial positioning technology to track the user's head position and viewing angle. Denote the initial position as (x0, y0, z0), and the real-time position is (x, y, z). Then the viewing angle position update formula is: Δx = x - x0, Δy = y - y0, Δz = z - z0, and adjust the model rendering viewing angle through the view matrix transformation to achieve the immersive viewing effect of the user in the virtual space.

[0016] Preferably, in step S2, the specific implementation of analyzing preferences through an improved recommendation algorithm to generate personalized vehicle type recommendations in real time is as follows:

[0017] S21. First, convert user preferences into a feature vector: V user =(S1, S2,... S m ), where S i represents the preference score of the user for the i-th type of vehicle. Vectorize the descriptions of each vehicle in the vehicle dataset to obtain: V model =(M1, M2,... M m ), where M i represents the feature description of the i-th vehicle model;

[0018] S22. Secondly, introduce weighted cosine similarity calculation. The weighted cosine similarity formula is: where ω i is the weight of each feature. Based on the similarity ranking, recommend the several vehicle models with the highest scores to the user, and dynamically adjust the recommended content according to the user's real-time interaction behavior;

[0019] S23. Finally, record the user's feedback on the recommended content, and fine-tune the preference feature vector so that the improved recommendation algorithm can quickly adapt to the user's latest interests. The preference update formula is: V user =αV user-old +(1 - α)V feedback , where α is the adjustment coefficient, V user-old is the original preference setting, and V feedback is the user feedback vector.

[0020] Preferably, in step S4, for the user's test drive, a highly immersive virtual reality experience system is constructed to simulate the driving experience. The specific implementation steps are as follows:

[0021] S41. First, provide an intuitive interface. The user can select different driving scenarios. According to the selected scenario, the system loads the corresponding 3D environment model in real time, and uses scene rendering technology to make the environment have a real visual effect. Set specific environmental characteristics for each scenario, including weather, time, and road conditions, to enhance the immersion. The characteristics are represented as: where w i is the weather, t i is the time, and r i is the road condition;

[0022] S42. Secondly, support the user to use a steering wheel, pedals, or handle device for control, and collect the user's control input in real time. The input data is represented as: I = {a, s, t}, where a is acceleration, s is the steering angle, and t is braking;

[0023] S43. Finally, while the user is operating, the acceleration and steering effects are real-time feedback, including steering wheel vibration, seat movement, and visual special effects, to enhance the immersion.

[0024] Preferably, in step S5, the gaze tracking technology based on the clustering algorithm captures the specific areas that the user is interested in and generates a visual heat map to identify the specific operations of the user's preference focus as follows:

[0025] S51. First, the gaze trajectory data of the user in the virtual exhibition hall is obtained in real time through the VR device, including the viewing angle position, the stay duration, and the gaze path. The set of gaze data points is G = {g1, g2,... g n}, where g i contains the coordinates (x i , y i ) and the timestamp t i of this data point;

[0026] S52. Then, by calculating the data point density at each position, the high-frequency areas where the user's gaze stays are obtained. The stay points are calculated as: where δ(g i , g j ) represents the distance between the gaze points g i and g j . The larger the value, the longer the stay time and the higher the attention;

[0027] S53. Next, the density-based clustering algorithm is used to cluster the gaze data points in real time, and the areas where the user's gaze is concentrated are divided into several high-density hot spots. The clustering results are used to identify the attention hot spots in the virtual exhibition hall. The attention intensity H i of each clustering center c i is the number of gaze points it contains, and weights are assigned to each hot spot area, with the attention duration and gaze point density of the hot spot area as the core indicators: where ω j represents the time weight of the gaze point. The greater the area intensity, the higher the user's attention;

[0028] S54. Finally, based on the clustering results of the gaze hot spot areas, a visual heat map is generated through the interpolation algorithm, with the color depth representing the hot spot intensity, and the user's preference hot spots in the virtual exhibition hall are updated in real time. The gaze heat map formula is: where (x i , y i ) represents the clustering center position, and σ controls the diffusion range.

[0029] Compared with the existing technology, the advantages and positive effects of the present invention are that it improves the realism of virtual displays through high-precision 3D modeling and physical lighting rendering technology; the intelligent recommendation algorithm analyzes user preferences in real time, provides personalized vehicle model recommendations, and enhances interactivity; supports immersive test drive simulation, combined with diverse driving scenarios and real-time feedback, to enrich the user experience; and eye tracking technology generates visual heat maps, accurately captures user attention areas, and provides effective data support for subsequent marketing and exhibition planning. DETAILED DESCRIPTION

[0030] In order to more clearly understand the above-mentioned objects, features and advantages of the present invention, the present invention is further described below in conjunction with the embodiments. It should be noted that, in the absence of conflict, the embodiments of the present application and the features therein can be combined with each other.

[0031] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways than those described herein. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0032] In this embodiment, with the development of virtual reality technology, auto shows are gradually introducing immersive interactive experiences. However, existing display methods suffer from insufficient immersion, lack of personalized recommendations, and limited interactivity, making it difficult to meet users' growing demand for in-depth experiences. This patent proposes a virtual reality-based interactive experience method for auto shows. Combining technologies such as 3D modeling, personalized recommendations, and virtual test drive simulation, it provides an immersive vehicle display and configuration experience. It also uses eye tracking to analyze user preferences, providing data support for subsequent marketing.

[0033] In order to enhance the user's immersive experience, we first use 3D modeling technology to build a model of the virtual showroom and the displayed models. First, we use refined modeling technology to build a 3D model of the showroom environment and the vehicle. The model is divided into an exterior geometry model and an internal structure model. The overall model M is expressed as: M = M ext +M int , where M ext Represents the vehicle appearance model, M int Represents the interior model of the vehicle; then, detail enhancement technology is applied to the vehicle model M, and polygon details are increased by the surface subdivision method to make the surface smoother; then, a high-resolution map is used to texture map the surface of the model, and the map is represented by T ext and T int , corresponding to the appearance and internal structure respectively, the texture of the model M after mapping is: M=M ext ·T ext +M int ·T intAnd use a physically based lighting model to enhance realism. The lighting equation is: L(o) = k d ·(N·L i )·I i +k s ·(R·V) n ·I i , where L(o) is the output brightness, k d , k s are the diffuse and specular reflection coefficients respectively, N is the normal vector, L i is the incident light, R is the reflection vector, V is the viewing vector, I i is the incident light intensity, and n is the specular coefficient; Additionally, real-time rendering is achieved based on deferred shading technology. Geometric information, normal vectors, textures, and other data are first rendered to the buffer, and then lighting calculations are performed to optimize the rendering efficiency. The target real-time rendering frame rate is F ≥ 90 FPS to meet the smoothness of the immersive experience; Finally, spatial positioning technology is used to track the user's head position and viewing angle. Denote the initial position as (x0, y0, z0), and the real-time position is (x, y, z). Then the viewing angle position update formula is: Δx = x - x0, Δy = y - y0, Δz = z - z0, and the model presentation viewing angle is adjusted through the view matrix transformation to achieve the immersive viewing effect of the user in the virtual space. Through high-precision modeling and physical lighting, the realism of the vehicle is enhanced. Deferred shading ensures a high frame rate for real-time rendering and improves the user's immersive experience.

[0034] Next, considering recording the user's browsing behavior to better generate personalized vehicle type recommendation services for users, the present invention adopts an improved recommendation algorithm. First, the user preferences are converted into a feature vector: V user =(S1, S2,... S m ), where S i represents the preference score of the user for the i-th type of vehicle. The descriptions of each vehicle in the vehicle dataset are vectorized to obtain: V model =(M1, M2,... M m ), where M i represents the feature description of the i-th vehicle model; Secondly, the weighted cosine similarity calculation is introduced. The weighted cosine similarity formula is: where ω i is the weight of each feature. Based on the similarity ranking, the several vehicle models with the highest scores are recommended to the user, and the recommended content is dynamically adjusted according to the user's real-time interaction behavior; Finally, the user's feedback on the recommended content is recorded, and the preference feature vector is finely tuned so that the improved recommendation algorithm can quickly adapt to the user's latest interests. The preference update formula is: V user =αV user-old +(1 - α)V feedback , where α is the adjustment coefficient, V user-oldFor the original preference setting, V feedback Is the user feedback vector. Analyze user preferences, recommend models in real time, dynamically adjust content, optimize the personalized experience, and enhance user engagement and the interactivity of the display.

[0035] After the user selects a model, they can configure it. The system provides color, wheel, and seat configuration options and renders them in real time for display. Then the user can choose to test drive. The system loads different driving scenarios, simulates the control experience, and provides feedback on acceleration and steering effects, enabling the user to experience the vehicle's performance. First, an intuitive interface is provided where the user can select different driving scenarios. Based on the selected scenario, the system loads the corresponding 3D environment model in real time and uses scene rendering technology to make the environment have a realistic visual effect. Specific environmental characteristics are set for each scenario, including weather, time, and road conditions, to enhance the immersion. The characteristics are represented as: Where w i Is the weather, t i Is the time, r i Is the road condition; Second, it supports the user to use a steering wheel, pedals, or a handle device for control, and real-time collects the user's control input. The input data is represented as: I = {a, s, t}, where a is acceleration, s is the steering angle, and t is braking; Finally, while the user is controlling, real-time feedback on acceleration and steering effects is provided, including steering wheel vibration, seat movement, and visual effects, to enhance the immersion. This operation supports diverse driving scenarios and real-time feedback, simulates the real driving experience through device interaction, showcases the vehicle's performance, and enhances the fun of the test drive.

[0036] Finally, to optimize user preference recognition, the present invention tracks and analyzes the line-of-sight trajectory based on a clustering algorithm, captures the specific areas that the user is interested in, and generates a visual heat map to identify the key points of user preferences. First, real-time line-of-sight trajectory data of the user in the virtual exhibition hall is obtained through a VR device, including the viewing angle position, dwell time, and line-of-sight path. The set of line-of-sight data points is G = {g1, g2,... g n}}, where g i Contains the coordinates (x i , y i ) and the timestamp t i ; Then, by calculating the data point density at each position, the high-frequency areas where the user's line of sight stays are obtained. The stay points are calculated as: Where δ(g i , g j ) Represents the distance between the line-of-sight points g i And g j . The larger the value, the longer the stay time and the higher the attention; Then, a density-based clustering algorithm is used to cluster the line-of-sight data points in real time, and the areas where the user's line of sight is concentrated are divided into several high-density hot spots. The clustering results are used to identify the attention hot spots in the virtual exhibition hall. The c of each clustering centeri Attention intensity H i It includes the number of line-of-sight points, and assigns weights to each hot spot area, with the attention duration and line-of-sight point density of the hot spot area as the core indicators: where ω j represents the time weight of the line-of-sight point. The greater the area intensity, the higher the user's attention. Finally, based on the clustering results of the line-of-sight hot spot area, a visual heat map is generated through an interpolation algorithm, with the color depth representing the hot spot intensity, and the preferred hot spots of the user in the virtual exhibition hall are updated in real time. The formula for the line-of-sight heat map is: where (x i , y i ) represents the position of the clustering center, and σ controls the range of diffusion. A visual heat map is generated through line-of-sight trajectory analysis to identify the user's preferred area, providing data support for subsequent marketing, improving the efficiency of exhibition data collection, and enhancing the response ability of exhibition activities to user needs.

[0037] The above description is only a preferred embodiment of the present invention, and it is not intended to limit the present invention in other forms. Any person skilled in the art may use the disclosed technical content to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as the technical solution content of the present invention is not departed from, any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A virtual reality-based interactive experience method for automobile exhibitions, characterized in that It includes the following steps: S1. First, use 3D modeling technology to construct 3D models of the virtual exhibition hall and the exhibited vehicle models in the hall, and optimize the models to enable users to immerse themselves in viewing the appearance and internal structure of the exhibited vehicle models; S2. Second, record the user's browsing behavior, analyze preferences through an improved recommendation algorithm, and generate personalized vehicle model recommendations in real time, which are displayed within the user's field of vision for further viewing; S3. Then, after the user selects a vehicle model, the system provides color, wheel hub, and seat configuration options and renders them in real time for display. The personalized configuration can be saved and used for sales communication; S4. When the user chooses to take a test drive, the system loads different driving scenarios, simulates the driving experience, and feedbacks the acceleration and steering effects to enable the user to feel the vehicle performance; S5. Finally, use the gaze tracking technology based on the clustering algorithm to capture the specific areas that the user focuses on and generate a visual heat map to identify the key points of user preferences, providing data support for subsequent marketing and exhibition planning; In step S5, the specific operation of the gaze tracking technology based on the clustering algorithm to capture the specific areas that the user focuses on and generate a visual heat map to identify the key points of user preferences is as follows: S51. First, the VR device is used to obtain the line-of-sight trajectory data of the user in the virtual exhibition hall in real time, including the viewing angle position, the staying duration, and the line-of-sight path. The set of line-of-sight data points is G = {g1, g2,... g n}, where g i contains the coordinates (x i , y i ) of this data point and the timestamp t i ; S52. Then, by calculating the data point density at each position, obtain the high-frequency area where the user's line of sight stays. The stay point is calculated as follows: where δ(g i , g j ) represents the distance between the line-of-sight points g i and g j . The larger the value, the longer the stay time and the higher the attention level. S53. Next, use the density-based clustering algorithm to cluster the line-of-sight data points in real time, divide the areas that the user's line of sight focuses on into several high-density hot spots, and use the clustering results to identify the attention hot spots in the virtual exhibition hall. The c of each clustering center i Attention intensity H i is the number of line-of-sight points it contains, and assign weights to each hot spot area, with the attention duration and line-of-sight point density of the hot spot area as the core indicators: where ω j represents the time weight of the line-of-sight point, and the greater the area intensity, the higher the user's attention; S54. Finally, based on the clustering results of the line-of-sight hot spot areas, a visual heat map is generated through an interpolation algorithm, with the intensity of the hot spots represented by the depth of color, and the preferred hot spots of the user in the virtual exhibition hall are updated in real time. The formula for the line-of-sight heat map is: where (x i , y i ) represents the position of the clustering center, and σ controls the range of diffusion.

2. The method for an interactive experience of an automobile exhibition based on virtual reality according to claim 1, wherein, In step S1, the specific implementation of using 3D modeling technology to construct 3D models of the virtual exhibition hall and the exhibited vehicle models in the hall and optimizing the models is as follows: S11. First, use the refined modeling technology to perform 3D modeling on the exhibition hall environment and the vehicle. The model is divided into an exterior geometric model and an internal structure model. The overall model M is expressed as: M = M ext + M int , where M ext represents the vehicle exterior model, and M int represents the vehicle interior model; S12. Then, apply the detail enhancement technology to the vehicle model M, and increase the polygon details through the surface subdivision method to make the surface smoother; S13. Then, texture mapping is performed on the model surface using high - resolution texture maps, and the texture maps are denoted as T ext and T int , corresponding to the appearance and internal structure respectively. After mapping, the texture of the model M is: M = M ext ·T ext +M int ·T int And a physically - based lighting model is used to enhance the realism. The lighting equation is: L(o) = k d ·(N·L i )·I i +k s ·(R·V) n ·I i , where L(o) is the output brightness, k d , k s are the diffuse and specular reflection coefficients respectively, N is the normal vector, L i is the incident light, R is the reflection vector, V is the viewing vector, I i is the incident light intensity, and n is the specular coefficient; S14. In addition, implement real-time rendering based on the deferred shading technology. Render the geometric information, normal vectors, and texture data to the buffer first, and then perform the lighting calculation to optimize the rendering efficiency. The target real-time rendering frame rate is F≥90FPS to meet the smoothness of the immersive experience; S15. Finally, use the spatial positioning technology to track the user's head position and viewing angle. Denote the initial position as (x0, y0, z0), and the real-time position is (x, y, z). Then the viewing angle position update formula is: Δx = x - x0, Δy = y - y0, Δz = z - z0, and adjust the model presentation viewing angle through the view matrix transformation to achieve the immersive viewing effect of the user in the virtual space.

3. The method for an interactive experience of an automobile exhibition based on virtual reality according to claim 1, wherein, The specific implementation of analyzing preferences through an improved recommendation algorithm and generating personalized vehicle model recommendations in real time in step S2 is as follows: S21. First, convert the user preference into a feature vector: V user =(S1, S2,... S m ), where S i represents the preference score of the user for the i-th type of vehicle model. Vectorize the descriptions of each vehicle model in the vehicle model dataset to obtain: V model =(M1, M2,... M m ), where M i represents the feature description of the i-th vehicle model; S22. Next, introduce the calculation of weighted cosine similarity. The formula for weighted cosine similarity is as follows: where ω i is the weight of each feature. Based on the similarity ranking, recommend several models with the highest scores to the user, and dynamically adjust the recommended content according to the user's real-time interaction behavior; S23. Finally, record the user's feedback on the recommended content and fine-tune the preference feature vector so that the improved recommendation algorithm can quickly adapt to the user's latest interests. The preference update formula is: V user = αV user-old + (1 - α)V feedback , where α is the adjustment coefficient, V user-old is the original preference setting, and V feedback is the user feedback vector.

4. A method for automotive exhibition interactive experience based on virtual reality according to claim 1, characterized in that For the user's test drive in step S4, a highly immersive virtual reality experience system is constructed to simulate the driving experience. The specific implementation steps are as follows: S41. First, provide an intuitive interface where users can select different driving scenarios. According to the selected scenarios, the system will load the corresponding 3D environment models in real time and use scene rendering technology to endow the environment with a realistic visual effect. Specific environmental characteristics are set for each scenario, including weather, time, and road conditions, to enhance the immersion. The characteristics are represented as: where w i is the weather, t i is the time, and r i is the road condition; S42. Second, support the user to use a steering wheel, pedals, or handle device for control, and collect the user's control input in real time. The input data is expressed as: I = {a, s, t}, where a is acceleration, s is the steering angle, and t is braking; S43. Finally, while the user is controlling, feedback the acceleration and steering effects in real time, including steering wheel vibration, seat movement, and visual special effects, to enhance the immersion.

Images