Scenic spot recommendation system adopting virtual reality technology

Abstract

The invention discloses a tourist attraction recommendation system adopting a virtual reality technology, belongs to the field of image data processing, and is used for solving the problems of how to improve defects in image information processing and improve scene fidelity and how to increase a virtual reality scene interaction function and improve scene interactivity. The system comprises seven main modules including a data acquisition module, a virtual reality scene construction module, a user information acquisition module, a recommendation algorithm module, a virtual reality experience module, a feedback collection module and a recommendation adjustment module. According to the technical scheme, through integration of the above modules, in the image modeling unit, a finer three-dimensional modeling technology is combined with accurate geographic information fusion and rich text information embedding operation, so that real scenic spots in a virtual scene can be restored as much as possible in a one-grass-one-wood and one-tile manner, highly-vivid and immersive experience is provided for a user, and the user experience is improved. And the user can feel the real atmosphere and detail characteristics of the scenic spot like being personally on the scene.

Description

Technical Field

[0001] The present invention belongs to the field of image data processing. Specifically, it is a tourist attraction recommendation system using virtual reality technology. Background Art

[0002] A virtual reality tourism method (Application No. CN201810288714.5) in the background art discloses a virtual reality tourism method including the following steps: connecting a virtual reality device to a tourism service platform; inputting a scenic spot resource selection instruction through the virtual reality device, and the tourism service platform receiving and transmitting the corresponding scenic spot resource to the virtual reality device according to the scenic spot resource selection instruction, and the virtual reality device receiving and playing the scenic spot resource; inputting a tour guide demand service instruction through the virtual reality device, and the tourism service platform receiving and transmitting the corresponding tour guide resource to the virtual reality device according to the tour guide demand service instruction, and the virtual reality device receiving and playing the tour guide resource.

[0003] Although this technical solution can provide a virtual reality tourism experience, in constructing a virtual scene, the processing of some details is not perfect. For example, in image information processing, it is impossible to achieve fine integration of all-round and multi-angle images of scenic spots, resulting in a lack of vividness of the scene. Due to the lack of image information processing, the fusion of geographical information of scenic spots is not accurate enough, resulting in inaccurate presentation of topographic and geomorphic features; this may allow users to feel the immersion of virtual reality during the experience process, but there is still room for improvement in the closeness to the real tourism scene, affecting the accurate judgment of users on the real situation of scenic spots.

[0004] A real-time holographic virtual reality system for scenic spots based on cloud technology (Application No. CN201610084577.4) in the background art discloses the following: a real-time holographic virtual information capture subsystem for capturing real-time holographic virtual information including three-dimensional holographic images of scenic spots and scenic spot environment information (temperature, noise); a cloud service subsystem including cloud storage, cloud transmission, and cloud computing, for receiving the real-time holographic virtual information captured by the real-time holographic virtual information capture subsystem, performing relevant processing calculations, and then transmitting it to each real-time holographic virtual reproduction area; and a real-time holographic virtual reproduction subsystem responsible for reproducing the real-time holographic virtual information sent by the cloud service subsystem to synchronously reproduce the current situation of the scenic spot in the real-time holographic virtual reproduction area for tourists to travel in the reproduced virtual scenic spot.

[0005] This technical solution realizes the storage and processing of data with the help of cloud technology, and focuses on providing a real-time holographic virtual reality experience for scenic spots. It may focus more on real-time in data processing. For example, it can obtain various dynamic information of scenic spots in real time (such as the distribution of tourist flow, the impact of real-time weather conditions on the scenic view), and quickly integrate this information into the virtual reality scene and present it to users. However, due to the lack in image information processing, for the presentation of the details of scenic spots, due to the relatively high real-time requirements, the restoration degree of some static details (such as the fine texture of historical buildings, the microscopic features of cultural relics) may be relatively limited, and the holographic effect may be more reflected in the real-time dynamic display of the overall scene.

[0006] To sum up, although virtual reality technology is constantly developing, the fidelity of the virtual reality scenes constructed by some current tourist attraction recommendation systems still needs to be improved. For example, in terms of detail restoration, it is difficult to present some complex natural landscapes (such as the vegetation diversity in a real forest, the dynamic effect of water flow) or the fine texture and interior decoration of historical buildings completely realistically, resulting in a certain gap between the user experience and the real visit during the experience process, and the user cannot obtain an extreme immersive experience.

[0007] The interaction function of the virtual reality scenes constructed by some systems is relatively weak. Users can often only perform some simple operations, such as moving and viewing basic information. For more in-depth interaction requirements, such as participating in activities held at scenic spots in the virtual scene and having a more realistic dialogue interaction with virtual characters (such as tour guides, local residents), it is not done well enough, reducing the user's participation and the fun of the experience in the virtual scene.

[0008] Therefore, whether it is to improve the defects in image information processing, enhance the scene fidelity, or to increase the interaction function of the virtual reality scene and improve the scene interactivity. There is an urgent need to develop a recommendation system that uses virtual reality technology to construct virtual reality scenes for tourist attractions, including processing the image information of scenic spots to create a three-dimensional scene model and adding texture operations. Summary of the Invention

[0009] The purpose of the present invention is to provide a tourist attraction recommendation system using virtual reality technology that can not only improve the defects in image information processing, enhance the scene fidelity, but also increase the interaction function of the virtual reality scene and improve the scene interactivity.

[0010] To achieve the above technical purpose, the technical solution adopted by the present invention is as follows:

[0011] A data acquisition module, which is used to collect relevant data of multiple tourist attractions. The relevant data includes scenic spot image information, scenic spot geographical information, scenic spot introduction text information, and tourist evaluation information;

[0012] A virtual reality scene construction module, connected to the data collection module, is used to construct virtual reality scenes corresponding to each tourist attraction according to the collected data related to tourist attractions. Each virtual reality scene can simulate the real tour experience of tourists within the corresponding tourist attraction;

[0013] A user information acquisition module, used to acquire the personal information and travel preference information of users. The personal information includes age, gender, and geographical location, and the travel preference information includes travel purpose, type of attractions of interest, and expected travel duration;

[0014] A recommendation algorithm module, respectively connected to the virtual reality scene construction module and the user information acquisition module, is used to screen and match according to the personal information and travel preference information of users among the constructed virtual reality scenes of each tourist attraction, and generate a preliminary recommended attractions list;

[0015] A virtual reality experience module, connected to the recommendation algorithm module, is used to provide users with virtual reality scene experiences corresponding to each tourist attraction in the preliminary recommended attractions list, so that users can intuitively feel the actual situation of each recommended attraction;

[0016] A feedback collection module, connected to the virtual reality experience module, is used to collect the feedback information of users during the process of experiencing the virtual reality scene. The feedback information includes the degree of preference of users for each attraction, the specific areas of attractions of interest, and the evaluation of the virtual reality experience effect;

[0017] A recommendation adjustment module, respectively connected to the feedback collection module and the recommendation algorithm module, is used to adjust and optimize the preliminary recommended attractions list according to the feedback information of users, and generate a final recommended attractions list to be provided to users.

[0018] The invention adopting the above technical solution can achieve the following effects.

[0019] First, personalized and accurate recommendation,

[0020] The system obtains in detail the personal information of users, such as age, gender, and geographical location, as well as travel preference information, including travel purpose, type of attractions of interest, and expected travel duration. With the help of the advanced recommendation algorithm module, these factors are accurately matched with the rich tourist attraction database. Whether it is an older tourist who likes historical and cultural relics or a young traveler who pursues excitement and adventure, they can obtain attractions recommendations that meet their own needs. This personalized and accurate recommendation greatly saves the time for users to screen attractions, improves the efficiency of travel planning, and enables users to lock in their favorite travel destinations more quickly.

[0021] Second, immersive experience and feel in advance,

[0022] The virtual reality scene construction module creates a realistic virtual reality scene for each tourist attraction. Without having to physically visit the attraction, users can experience the charm of the attraction as if they were there by operating the device through a head-mounted display and a controller. Every detail, from the bricks and tiles of ancient buildings to the landscape paintings of natural scenery, is vivid. This immersive experience enables tourists to have an in-depth understanding of the attraction before departure, better plan their tour routes, identify the areas they are truly interested in, avoid missing out on wonderful sights due to lack of understanding after arriving at the attraction, and at the same time increases the sense of anticipation for the trip.

[0023] III. Comprehensive and realistic presentation of attractions,

[0024] The data collection module comprehensively collects relevant data of tourist attractions, including scenic spot image information, geographical information, introduction text information, and tourist evaluation information. In the virtual reality scene, these information are organically integrated. The geographical location relationship and topographical features of the scenic spot are accurately presented through the integration of geographical information. The scenic spot introduction text is displayed in voice or text form in a timely manner, and tourist evaluations are also provided for users to refer to in a visual way. Users can understand the true features of a scenic spot from multiple perspectives and comprehensively, rather than just seeing some one-sided promotional pictures, so as to make more accurate travel decisions.

[0025] IV. Real-time feedback for optimized recommendations,

[0026] The collaborative work of the feedback collection module and the recommendation adjustment module enables the recommendation system to optimize the recommended scenic spot list in real time according to the feedback information of users during the virtual reality experience process. If users are not satisfied with the virtual reality experience of a certain scenic spot, or find that the actual situation does not meet their expectations, the system will adjust the recommendation in a timely manner and push scenic spots that are more in line with users' preferences to users. This dynamic optimization mechanism ensures that the recommendation results always closely match the real needs of users and improves users' satisfaction with the recommendation system.

[0027] V. Social interaction to expand the experience,

[0028] The social sharing module provides users with the function of sharing their virtual reality experience feelings and recommended scenic spot information through social network platforms. Users can share their discoveries with friends and family, attracting more people's attention to tourist attractions. At the same time, others can also initially feel the charm of the scenic spot through the virtual reality scene link or video clip attached to the shared content, further expanding the publicity scope of tourist attractions and promoting the spread of tourism culture.

[0029] VI. Improving the quality of travel decisions,

[0030] Combining the advantages in all aspects above, the tourist attraction recommendation system of the present invention can help users understand each attraction more comprehensively and deeply during the tourism planning stage. Instead of choosing attractions solely based on vague impressions or limited promotional materials, users make decisions based on real and immersive experiences and accurate recommendations. This undoubtedly greatly improves the quality of tourism decisions, making each trip more likely to become an unforgettable and wonderful experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The present invention can be further illustrated by the non-limiting embodiments given in the accompanying drawings;

[0032] Figure 1 It is a schematic diagram of the main functional modules of the tourist attraction recommendation system of the present invention;

[0033] Figure 2 It is a flowchart of the main functional modules of the tourist attraction recommendation system of the present invention;

[0034] Figure 3 It is a partial flowchart of the data acquisition module of the tourist attraction recommendation system of the present invention;

[0035] Figure 4 It is a schematic diagram of the steps for constructing a three-dimensional scene model in the image modeling unit of the tourist attraction recommendation system of the present invention;

[0036] Figure 5 It is a schematic diagram of the virtual reality experience module of the tourist attraction recommendation system of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0037] In order to enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0038] Embodiment 1

[0039] As Figure 1 、 Figure 2 、 Figure 3 、 Figure 4 shown, a tourist attraction recommendation system using virtual reality technology includes:

[0040] A data acquisition module, used to collect relevant data of multiple tourist attractions, and the relevant data includes scenic spot image information, scenic spot geographical information, scenic spot introduction text information, and tourist evaluation information;

[0041] A virtual reality scene construction module, connected to the data acquisition module, used to construct a virtual reality scene corresponding to each tourist attraction according to the collected relevant data of the tourist attractions, and each virtual reality scene can simulate the real tour experience of tourists in the corresponding tourist attraction;

[0042] A user information acquisition module, which is used to acquire the personal information and travel preference information of the user. The personal information includes age, gender, and geographical location. The travel preference information includes travel purpose, type of scenic spots of interest, and expected travel duration;

[0043] A recommendation algorithm module, which is respectively connected to the virtual reality scene construction module and the user information acquisition module. It is used to screen and match in the constructed virtual reality scenes of each tourist attraction according to the personal information and travel preference information of the user, and generate a preliminary recommended scenic spot list;

[0044] A virtual reality experience module, which is connected to the recommendation algorithm module. It is used to provide the user with virtual reality scene experiences corresponding to each tourist attraction in the preliminary recommended scenic spot list, so that the user can intuitively feel the actual situation of each recommended scenic spot;

[0045] A feedback collection module, which is connected to the virtual reality experience module. It is used to collect the feedback information of the user during the process of experiencing the virtual reality scene. The feedback information includes the degree of preference of the user for each scenic spot, the specific area of the scenic spot that the user is interested in, and the evaluation of the virtual reality experience effect;

[0046] A recommendation adjustment module, which is respectively connected to the feedback collection module and the recommendation algorithm module. It is used to adjust and optimize the preliminary recommended scenic spot list according to the feedback information of the user, and generate a final recommended scenic spot list to be provided to the user.

[0047] First, through the recommendation algorithm module and the recommendation adjustment module, the present invention comprehensively considers various aspects of user information and feedback, and uses complex image processing and recommendation algorithms to accurately match user needs and continuously optimize the recommendation results.

[0048] Second, through the data acquisition module, the virtual reality scene construction module, and the feedback collection module, the present invention deeply integrates image processing into the complete tourist attraction recommendation system process from data acquisition, scene construction to recommendation and feedback adjustment.

[0049] Third, through the virtual reality scene construction module and the virtual reality experience module, the present invention constructs a virtual reality scene that allows users to have an immersive experience. One of the cores of the present invention is to create a realistic and interactive virtual reality scene for recommendation by means of perfect image processing. In particular, for the screening of images and the adjustment of image scenes based on user feedback.

[0050] Fourth, through the integration of the above modules, in the image modeling unit, using more refined three-dimensional modeling technology combined with accurate geographical information fusion and rich text information embedding operations, every plant and tree, every brick and tile in the virtual scene are restored as realistically as possible to the real scenic spots, providing users with a highly realistic and immersive experience, making users feel as if they are on the spot and experiencing the real atmosphere and detailed features of the scenic spots.

[0051] In a preferred embodiment, the data acquisition module is further configured to update the relevant data of the tourist attractions in real time, so as to ensure that the constructed virtual reality scene can reflect the latest status of the tourist attractions.

[0052] In a preferred embodiment, the virtual reality scene construction module includes:

[0053] An image modeling unit, configured to construct a three-dimensional scene model of the tourist attraction by using three-dimensional modeling technology according to the scenic spot image information;

[0054] A geographic information fusion unit, configured to fuse the scenic spot geographic information into the three-dimensional scene model to accurately present the geographical location relationship and topographic and geomorphic features of the scenic spot;

[0055] A text information embedding unit, configured to embed the scenic spot introduction text information in the corresponding position of the three-dimensional scene model in the form of voice or text display, so as to provide scenic spot introduction for users during the virtual reality experience;

[0056] A tourist evaluation display unit, configured to visually display the tourist evaluation information in the evaluation area of the three-dimensional scene model for users to refer to.

[0057] In a preferred embodiment, the recommendation algorithm module uses a combination of a content-based recommendation algorithm and a collaborative filtering recommendation algorithm for scenic spot screening and matching. Among them, the content-based recommendation algorithm screens according to the similarity between the attribute characteristics of the tourist attractions and the user's travel preference information, and the collaborative filtering recommendation algorithm screens according to the selection of tourist attractions by other users with similar travel preferences.

[0058] In a preferred embodiment, the virtual reality experience module includes:

[0059] A head-mounted display device, configured to provide an immersive virtual reality scene viewing experience for users. It has a high-resolution display screen and an accurate head tracking function to update the virtual reality scene perspective in real time as the user's head rotates;

[0060] A handle operation device, connected to the head-mounted display device, for users to perform interactive operations in the virtual reality scene, such as selecting a tour route, viewing detailed information of scenic spots, and switching the scene perspective.

[0061] In a preferred embodiment, the feedback collection module collects user feedback information in the following ways:

[0062] During the virtual reality experience process, a feedback interaction interface is set up. Users can use the handle operation device to input the preference score for each scenic spot, select the specific area of the scenic spot they are interested in, and input a text evaluation of the virtual reality experience effect on this interface;

[0063] Real-time monitor the user's behavior data in the virtual reality scenario, such as stay time, browsing path, operation frequency, and analyze the behavior data to obtain the user's potential interest tendency in each scenic spot as part of the feedback information.

[0064] In the preferred solution, the specific way for the recommended adjustment module to adjust and optimize the preliminary recommended scenic spot list according to the user's feedback information is as follows:

[0065] If the user's preference score for a certain scenic spot is lower than the set threshold and the reason for expressing disinterest in the feedback is clear, such as the scenic spot type does not match or the experience effect is not good, then remove the scenic spot from the preliminary recommended scenic spot list;

[0066] According to the degree of interest of the user in the specific areas of each scenic spot in the virtual reality scenario, adjust the ranking of the corresponding scenic spots in the recommended list, and rank the scenic spots that the user is more interested in at the front;

[0067] Combined with the user's evaluation of the virtual reality experience effect, conduct targeted optimization and improvement on the subsequent constructed virtual reality scenario.

[0068] In the preferred solution, it also includes a scenic spot popularity analysis module, which is connected to the data collection module and the recommendation algorithm module, and is used to statistically analyze the popularity of each tourist scenic spot according to the collected tourist evaluation information and the number of visitors to each tourist scenic spot, and provide the scenic spot popularity information as an auxiliary factor to the recommendation algorithm module, so as to comprehensively consider the scenic spot popularity when generating the recommended scenic spot list.

[0069] In the preferred solution, the user information acquisition module obtains the user's travel preference information in the following ways:

[0070] Provide a travel preference questionnaire interface that the user can fill in. The questionnaire content includes travel purpose, scenic spot types of interest, and expected travel duration options. The user submits travel preference information by selecting or filling in relevant content;

[0071] Analyze the user's historical browsing records and reservation records on past travel-related platforms, and extract the information related to travel preferences therein as a supplement to the user's travel preference information.

[0072] In the preferred solution, the system also includes a social sharing module, which is connected to the virtual reality experience module, and is used to provide the user with the function of sharing the experience feelings and recommended scenic spot information through social network platforms after the user completes the virtual reality scenario experience, and the shared content can be attached with the virtual reality scenario link or video clip of the corresponding scenic spot for other users to view and experience.

[0073] In the preferred solution, for the data collection module, the steps for collecting scenic spot image information are as follows,

[0074] S101 determines the collection target and plan,

[0075] Clarify the scope of scenic spots to be collected, and determine the specific list of tourist attractions that need to collect image information based on the target coverage of the tourist attraction recommendation system. This may involve scenic spots in different regions and of different types (such as natural landscapes, historical and cultural relics).

[0076] Develop a collection plan and make a detailed image collection plan for each selected tourist attraction. Considerations include different areas of the attraction (such as entrances, main attractions, and special areas), different time periods (such as daytime, nighttime, and different seasons), and collection arrangements under different weather conditions to obtain comprehensive and diverse image data.

[0077] S102 Collection equipment preparation,

[0078] Choose UAV as the image acquisition device.

[0079] Drones can overlook the entire scenic spot from the air, obtain panoramic images and unique perspectives, which are very helpful in showing the overall picture, topography and layout of the scenic spot.

[0080] Among them, for drones: set the flight altitude, shooting angle, shooting mode (such as panoramic shooting, fixed-point shooting) and image resolution parameters to ensure that clear, comprehensive and distinctive aerial images can be obtained.

[0081] S103 field collection operation,

[0082] Use drones to take aerial photos.

[0083] Before takeoff, at a suitable takeoff location (usually an open, obstacle-free area that complies with drone flight regulations), perform various pre-takeoff checks on the drone, including battery power, propeller installation, and remote control connection, to ensure that the drone can fly normally.

[0084] Plan the flight route. According to the layout of the scenic spot and the image content to be collected, plan the flight route on the drone's flight control software. The flight route should cover the main area of ​​the scenic spot and try to obtain panoramic images and images of the scenic spot from different heights and angles.

[0085] Aerial photography: control the drone to fly and take pictures according to the planned flight route. During the flight, pay attention to keep a safe distance from obstacles and adjust the flight altitude, shooting angle and shooting mode as needed to obtain high-quality aerial images.

[0086] Landing and data export: After the shooting is completed, control the drone to land safely, and then export the captured image data from the drone's storage device to a specified storage medium (such as a computer hard drive or a portable hard drive), and also record relevant information about the shooting, such as flight time, flight route, and shooting altitude.

[0087] S104 Supplementary collection of network resources

[0088] Search for public image resources related to tourist attractions on the Internet as supplements. These network resources come from the official websites of scenic spots, travel forums, and social media platforms. When using network resources, it is necessary to pay attention to verifying the copyright situation of the images to ensure legal acquisition and use. For the images obtained from the Internet, relevant information such as the source website and the release time should also be recorded for subsequent management.

[0089] In the preferred solution, in S102 equipment preparation for collection, the relationship between flight altitude and shooting range

[0090] Assume that the camera lens angle of view of the drone is θ, using a horizontal view angle, and the flight altitude is h. Then, in an ideal situation, the radius of the ground range that the drone can shoot in the horizontal direction can be calculated by the following formula:

[0091]

[0092] For example, if the horizontal view angle θ of the drone lens is 60° and the flight altitude h is 100 meters, then the radius r of the ground range that can be shot in the horizontal direction is

[0093]

[0094] That is, in the horizontal direction, a circular ground range with a radius of 57.7 meters centered directly below the drone can be shot.

[0095] In the preferred solution, in the image modeling unit, for constructing a three-dimensional scene model of a tourist attraction using three-dimensional modeling technology based on the scenic spot image information, the following steps are included:

[0096] S201 Image collection

[0097] Collect image materials of tourist attractions from multiple data sources. These data sources include but are not limited to high-definition photos taken by professional photographers, pictures uploaded by travel enthusiasts, promotional pictures provided by scenic spot officials, and aerial views and panoramic views taken by drone equipment. Ensure that the collected images can cover different regions, different angles, and different seasons of the scenic spot to obtain comprehensive visual information.

[0098] Perform preliminary screening and sorting on the collected images, remove the blurred, highly repetitive images that cannot be used for modeling, and retain the representative and high-quality images as the basic materials for subsequent modeling.

[0099] S202 Feature extraction

[0100] For each of the screened images, use feature extraction algorithms in computer vision technology, such as SIFT (Scale-Invariant Feature Transform), SURF (Speeded-Up Robust Features), or ORB (Oriented FAST and Rotated BRIEF) algorithms, to extract the key feature points and their descriptions in the images. These feature points contain the position information of the unique visual characteristics in the image, such as edges, corners, and texture change parts, while the description is the quantitative representation of the features of these feature points, which is used for subsequent feature matching and 3D reconstruction.

[0101] Store the feature points and descriptions of all the images extracted to form a feature database for convenient query and use in subsequent steps.

[0102] S203 Image registration

[0103] Based on the previously extracted feature points and descriptions, find the corresponding relationships between different images through feature matching algorithms. Commonly used feature matching algorithms include the brute-force matching algorithm based on nearest neighbor search, and the fast matching algorithms based on KD-tree (K-Dimensional Tree) or FLANN (Fast Library for Approximate Nearest Neighbors). Through these algorithms, find the same or similar feature points in two or more images taken from different perspectives, so as to determine the relative position and pose relationship between these images.

[0104] Perform further optimization processing on the successfully matched image pairs, such as using the Random Sample Consensus (RANSAC) algorithm to remove the feature points with incorrect matches and improve the accuracy of image registration. After optimization, obtain a set of accurately registered image sequences, which will provide the basic data for subsequent 3D reconstruction.

[0105] S204 3D reconstruction

[0106] Based on the accurately registered image sequence, a 3D reconstruction algorithm based on multi-view geometry, such as Triangulation, is used to calculate the 3D coordinates of each feature point in the scene. Specifically, for each set of feature points that are successfully matched in different images, using the internal parameters of the camera (such as focal length, principal point position) and the known relative external parameters (such as rotation and translation relationships) when these images were taken, the coordinate position of this feature point in 3D space is calculated through the principle of triangulation.

[0107] As the 3D coordinates of more and more feature points are calculated, a sparse 3D point cloud model of the entire tourist attraction is gradually constructed. This point cloud model initially presents the 3D spatial structure of the attraction, but it is still relatively rough and consists only of discrete points.

[0108] S205 Point cloud optimization and densification,

[0109] The generated sparse 3D point cloud model is optimized to remove the outliers caused by noise and incorrect matching. Statistical filtering and radius filtering methods are used to identify and remove those outliers that are significantly inconsistent with the surrounding points according to the distribution law and distance relationship of the points in the point cloud, improving the quality of the point cloud model.

[0110] Through some point cloud densification techniques, such as the multi-view stereo (MVS) algorithm based on stereo vision and the point cloud generation algorithm based on deep learning, the sparse point cloud model is transformed into a dense point cloud model. The dense point cloud model contains more points and can present the surface shape and detailed features of the tourist attraction more delicately, providing a better basis for subsequent mesh generation.

[0111] S206 Mesh generation,

[0112] Based on the optimized dense point cloud model, a suitable mesh generation algorithm, such as Poisson surface reconstruction algorithm, moving least squares (MLS), is used to convert the point cloud into a 3D model in the form of a triangular mesh or polygon mesh. During the mesh generation process, according to the density and shape characteristics of the point cloud, the mesh cells are reasonably divided to ensure that the generated mesh model can accurately reflect the shape of the attraction and has a good topological structure, facilitating subsequent rendering and interactive operations.

[0113] The generated mesh model is further refined and optimized, such as through mesh smoothing and simplification operations, to remove the jagged flaws on the mesh surface, improving the smoothness and overall quality of the mesh model and making it more in line with the visual effect of the real scene.

[0114] S207 Texture mapping,

[0115] Select appropriate images from the initially captured high-quality image materials as texture maps, and accurately map the selected images onto the surface of the 3D model according to the mesh structure of the 3D model and the corresponding real-scene positions of each part. During the texture mapping process, ensure that the scale, orientation, and position of the texture image match the real scene to create a realistic visual effect.

[0116] Check and adjust the 3D model after texture mapping, handle possible texture stretching and distortion problems, and optimize the texture mapping effect by adjusting texture coordinates and reselecting appropriate texture images, ultimately constructing a complete 3D scene model of the tourist attraction.

[0117] In the preferred solution, in S201 image acquisition, the captured images are preliminarily screened and sorted through a preset threshold T V This preset threshold T V can be determined through experiments or experience. For example, for images that are clearly suitable for modeling, the variance is usually greater than a certain value. Specifically, for a grayscale image I(x,y), if it is a color image, it needs to be first converted to a grayscale image,

[0118] I(x,y) = 0.299R(x,y) + 0.587G(x,y) + 0.114B(x,y);

[0119] where R(x,y), G(x,y), and B(x,y) are the red, green, and blue channel values of the color image at this point, respectively.

[0120] First, calculate the gradient magnitude G(x,y) of the image. The Sobel operator can be used to calculate the gradients in the horizontal and vertical directions:

[0121] Horizontal direction gradient G x (x,y):

[0122] G x (x,y) = I(x + 1,y) - I(x - 1,y);

[0123] Vertical direction gradient G y (x,y):

[0124] G y (x,y) = I(x,y + 1) - I(x,y - 1);

[0125] Then the gradient magnitude G(x,y) is:

[0126]

[0127] Then calculate the variance V of the gradient magnitude of the entire image:

[0128]

[0129] where M and N are the number of rows and columns of the image respectively, is the average value of the image gradient magnitude, and the calculation formula is:

[0130]

[0131] When the variance V of the entire image gradient magnitude is less than the preset threshold T V , it can be considered that the image is blurred and can be removed.

[0132] Example 2,

[0133] such as Figure 1 , Figure 2 , Figure 3 , Figure 4 , Figure 5 shown, based on the technical solution of the above Example 1, we can also add a virtual reality experience module to build a virtual reality scene that allows users to have an immersive experience. One of the cores of the present invention is to create a realistic and interactive virtual reality scene for recommendation by means of perfect image processing. In the virtual reality experience module, the following modules are included,

[0134] S301 Enrich the interactive operation device and function module,

[0135] Optimize the handle operation,

[0136] Add function buttons. Add more function buttons to the existing handle operation device, and each button can correspond to different interactive actions. For example, in addition to the common direction keys and confirmation keys, set buttons specifically for triggering scenic spots or events. For example, pressing a certain button can start a virtual historical and cultural performance (for historical scenic spots), or trigger a special weather effect of a natural scenic spot (such as the appearance of a rainbow at a waterfall scenic spot).

[0137] Enhance the tactile feedback. Add a more delicate tactile feedback mechanism to the handle. When the user operates in the virtual scene, such as touching virtual objects (such as the walls of ancient buildings, the branches of trees), the handle can give corresponding vibration feedback according to the material characteristics of the touched object, allowing the user to more truly feel the texture of the objects in the virtual scene and enhancing the realism of the interaction.

[0138] Introduce a gesture recognition device,

[0139] Install a gesture recognition sensor. Equip the virtual reality experience area with a gesture recognition sensor, such as a camera-based visual gesture recognition system or an infrared gesture recognition device. Users do not need to hold an additional device and can interact with the virtual scene only through gesture actions. For example, waving the hand can switch the perspective of the virtual scene, and making a fist can grab virtual items (such as picking up leaves on the ground or picking up souvenirs at scenic spots). Different gesture combinations can correspond to different interaction instructions, greatly expanding the interaction methods.

[0140] Gesture action customization allows users to customize some gesture actions and their corresponding interaction functions according to their own habits and preferences. In this way, users can more flexibly control the interaction process and improve the convenience and personalization of the interaction.

[0141] S302 Expand the interaction elements and behavior modules in the virtual scene.

[0142] Character interaction

[0143] Add a virtual tour guide. Set up a virtual tour guide role in the virtual scene, which has an intelligent voice dialogue function and can answer various questions of users about scenic spots in real time, such as the historical background, architectural features, and tour route suggestions of the scenic spots. Users can interact with the virtual tour guide through voice or the above-mentioned handles and gestures, just like having a professional tour guide accompany them during a real tour.

[0144] Simulate local residents. For some scenic spots with cultural characteristics, simulate the activities of local residents in the virtual scene. Users can have simple communication and interaction with these virtual residents to understand local customs and folk stories. For example, in the virtual scene of an ancient town scenic spot, users can talk to virtual local residents and ask about the production methods of local special foods or the celebration methods of traditional festivals.

[0145] Object interaction

[0146] Increase the number of operable virtual objects. In addition to simply viewing the information of virtual objects, make more objects operable in the virtual scene. For example, in the virtual scene of a museum scenic spot, users can not only view cultural relics up close, but also operate the display cabinets of cultural relics (virtual operation, simulating the opening action in the real scene) through the handle or gestures to view the back or internal details of the cultural relics; in the garden scenic spot, users can move virtual flower pots and prune virtual branches, increasing the user's sense of participation and control in the virtual scene.

[0147] The state of the object changes, enabling the objects in the virtual scene to change their states according to the user's interaction behavior. For example, in a virtual bonfire party scene (a special activity scene set for some natural scenic spots), if the user approaches the bonfire and makes an action of adding firewood (simulated by a handle or gesture operation), the bonfire will burn more vigorously, and at the same time, the surrounding light and shadow effects will also change accordingly, allowing the user to feel the actual impact of their actions on the virtual scene.

[0148] S303 Strengthen the feedback mechanism module of scene interaction,

[0149] Immediate visual feedback,

[0150] Visualize the operation result. When the user performs any interaction operation in the virtual scene, the operation result is immediately presented with an intuitive visual effect. For example, if the user presses a certain key to trigger the light show effect of a scenic spot, gorgeous light changes should quickly appear in the virtual scene, and the color and blinking frequency of the lights can be adjusted accordingly according to the user's operation parameters (such as the duration and intensity of the key press, if applicable), allowing the user to clearly see the changes brought about by their operations.

[0151] Scene dynamic response. Other elements in the virtual scene should also make dynamic responses according to the user's interaction behavior. For example, when the user runs quickly in a virtual forest scenic spot (simulating the running action through a handle or gesture), the branches and leaves of the surrounding trees should sway with the wind, and small animals will be frightened and flee, creating a real and dynamic interaction environment.

[0152] Enhance audio feedback,

[0153] Match sound effects. Equip each interaction operation with corresponding audio feedback, and the sound effects should match the virtual scene and the operation content. For example, when the user touches virtual running water (such as a waterfall or a stream), they will hear a realistic sound of running water, and the volume of the running water will vary according to the position and intensity of the user's touch (if the touch intensity is simulated through a handle or gesture); when the user talks to a virtual tour guide, they can hear a clear voice response, and the volume and intonation of the voice will also be adjusted according to factors such as the distance (the concept of distance simulated in the virtual scene), enhancing the realism and immersion of the interaction.

[0154] Change of ambient sound effects. In addition to the sound feedback of the operation itself, the ambient sound effects of the entire virtual scene should also change according to the user's interaction behavior. For example, when the user enters a virtual cave scenic spot, the originally quiet ambient sound effects will become echoey as the user walks (simulating the walking action through a handle or gesture), and when the user makes a loud noise (such as shouting) in the cave, the echo effect will be more obvious, creating a real cave ambient sound effect experience.

[0155] The S304 realizes a multi-level interactive experience module.

[0156] Task-driven interaction

[0157] Set tour tasks, and set some tour tasks for users in the virtual scene. For example, in the virtual scene of a historical castle scenic spot, the user is required to find several cultural relics hidden in the castle within a specified time (search and discover through the handle and gesture methods). After completing the task, additional scenic areas can be unlocked or virtual souvenir rewards can be obtained. This task-driven interaction method can increase the user's participation and exploration desire in the virtual scene.

[0158] Plot experience tasks, and design plot experience tasks for some scenic spots with rich cultural connotations. For example, in a virtual ancient village scenic spot, the user can play the role of an ancient traveler and experience a series of events (such as helping villagers solve problems and participating in traditional festival celebrations) according to the preset plot clues (obtained by interacting with virtual tour guides and local residents), so as to experience the local cultural life and enhance the user's emotional experience and immersion in the virtual scene.

[0159] Social interaction expansion

[0160] Multi-person simultaneous experience, supporting multiple people to enter the same virtual scene for experience at the same time. Users can visit scenic spots, complete tasks, and communicate and interact with friends, family members or other tourists in the virtual scene. For example, in a virtual theme park scenic spot, multiple people can take a virtual roller coaster together and share their feelings with each other, increasing social fun and group experience.

[0161] Virtual community interaction, constructing a connection between the virtual community and the virtual scene. After experiencing the virtual scene, users can enter the virtual community and communicate and discuss with other users who have experienced the virtual scene of this scenic spot, sharing travel experiences and interactive experiences, further expanding the scope and depth of interaction and promoting the spread of tourism culture.

[0162] The above has introduced in detail a tourism scenic spot recommendation system using virtual reality technology provided by the present invention. The description of specific embodiments is only used to help understand the method and its core idea of the present invention. It should be noted that for those of ordinary skill in the art of this technology, without departing from the principle of the present invention, several improvements and modifications can be made to the present invention, and these improvements and modifications also fall within the protection scope of the claims of the present invention.

Claims

1. A tourist attraction recommendation system using virtual reality technology, characterized in that, Including: A data collection module, which is used to collect relevant data of multiple tourist attractions. The relevant data includes scenic spot image information, scenic spot geographical information, scenic spot introduction text information, and tourist evaluation information; A virtual reality scene construction module, which is connected to the data collection module and is used to construct a virtual reality scene corresponding to each tourist attraction according to the collected relevant data of the tourist attractions. Each virtual reality scene can simulate the real tour experience of tourists in the corresponding tourist attraction; A user information acquisition module, which is used to acquire the personal information and travel preference information of the user. The personal information includes age, gender, and geographical location. The travel preference information includes travel purpose, interested scenic spot type, and expected travel duration; A recommendation algorithm module, which is respectively connected to the virtual reality scene construction module and the user information acquisition module, and is used to screen and match according to the personal information and travel preference information in the constructed virtual reality scenes of each tourist attraction, and generate a preliminary recommended scenic spot list; A virtual reality experience module, which is connected to the recommendation algorithm module and is used to provide the user with a virtual reality scene experience corresponding to each tourist attraction in the preliminary recommended scenic spot list, so that the user can intuitively feel the actual situation of each recommended scenic spot; A feedback collection module, which is connected to the virtual reality experience module and is used to collect the feedback information of the user during the process of experiencing the virtual reality scene. The feedback information includes the user's preference degree for each scenic spot, the specific area of the scenic spot that the user is interested in, and the evaluation of the virtual reality experience effect; A recommendation adjustment module, which is respectively connected to the feedback collection module and the recommendation algorithm module, and is used to adjust and optimize the preliminary recommended scenic spot list according to the user's feedback information, and generate a final recommended scenic spot list to be provided to the user.

2. The tourism attraction recommendation system using virtual reality technology according to claim 1, wherein The virtual reality scene construction module includes: An image modeling unit, which is used to construct a three-dimensional scene model of a tourist attraction by using three-dimensional modeling technology according to the scenic spot image information; A geographical information fusion unit, which is used to fuse the scenic spot geographical information into the three-dimensional scene model to accurately present the geographical location relationship and topographic and geomorphic features of the scenic spot; A text information embedding unit, which is used to embed the scenic spot introduction text information in the corresponding position of the three-dimensional scene model in the form of voice or text display, so as to provide scenic spot introduction for the user during the virtual reality experience; A tourist evaluation display unit, which is used to visually display the tourist evaluation information in the evaluation area of the three-dimensional scene model for the user to refer to.

3. The tourism attraction recommendation system using virtual reality technology according to claim 1, characterized in that, The recommendation algorithm module uses a combination of a content-based recommendation algorithm and a collaborative filtering recommendation algorithm for scenic spot screening and matching. Among them, the content-based recommendation algorithm screens according to the similarity between the attribute characteristics of tourist attractions and the user's travel preference information, and the collaborative filtering recommendation algorithm screens according to the selection of tourist attractions by other users with similar travel preferences.

4. The tourism attraction recommendation system using virtual reality technology according to claim 1, wherein The virtual reality experience module includes: A head-mounted display device, which is used to provide the user with an immersive virtual reality scene viewing experience. It has a high-resolution display screen and an accurate head tracking function to realize real-time update of the virtual reality scene perspective as the user's head rotates; The handle operation device is connected to the head mounted display device and is used for the user to perform interactive operations in the virtual reality scene, such as selecting a tour route, viewing detailed information of scenic spots, and switching scene perspectives.

5. The tourism attraction recommendation system using virtual reality technology according to claim 1, characterized in that, The feedback collection module collects user feedback information in the following ways: During the VR experience, a feedback interaction interface is set up, on which users can use the handle to operate the device to input their preference ratings for each attraction, select specific areas of the attraction they are interested in, and input textual evaluations of the VR experience effects; Real-time monitoring of user behavior data in the virtual reality scene, such as dwell time, browsing path, and operation frequency, and analysis of the behavior data to obtain the user's potential interest in each attraction as part of the feedback information; The specific method in which the recommendation adjustment module adjusts and optimizes the preliminary recommended scenic spot list according to the user's feedback information is as follows: If the user's preference score for a certain attraction is lower than the set threshold and the reason for the lack of interest in the feedback is clear, such as the attraction type not matching or the experience effect is not good, the attraction will be removed from the preliminary recommended attractions list; According to the user's interest in the specific area of ​​each scenic spot in the virtual reality scene, the order of the corresponding scenic spots in the recommendation list is adjusted to put the scenic spots that the user is more interested in at the top; Based on users' evaluation of the virtual reality experience, targeted optimization and improvement are carried out on the subsequently constructed virtual reality scenes.

6. The tourism attraction recommendation system using virtual reality technology according to claim 1, wherein It also includes a scenic spot heat analysis module, which is connected to the data collection module and the recommendation algorithm module, and is used to statistically analyze the heat of each tourist attraction based on the collected tourist evaluation information and the number of visitors to each tourist attraction, and provide the scenic spot heat information as an auxiliary factor to the recommendation algorithm module, so as to comprehensively consider the heat of the scenic spot when generating a recommended scenic spot list; The user information acquisition module acquires the user's travel preference information in the following manner: Provide a travel preference questionnaire interface that users can fill out. The questionnaire content includes travel purpose, types of attractions of interest, and expected travel duration options. Users submit travel preference information by selecting or filling out relevant content; Analyze users' historical browsing records and booking records on past travel-related platforms, and extract information related to travel preferences as a supplement to users' travel preference information.

7. The tourism attraction recommendation system using virtual reality technology according to claim 1, characterized in that, The steps of collecting scenic spot image information in the data collection module are as follows: S101 determines the collection target and plan, Clarify the scope of scenic spots to be collected, and determine the list of specific tourist attractions that need to collect image information based on the target coverage of the tourist attraction recommendation system; S102 Collection equipment preparation, Choose drones as image acquisition devices. Drones can overlook the entire scenic spot from the air, obtain panoramic images and unique perspectives, and show the overall appearance, topography and layout of the scenic spot; S103 field collection operation, Plan the flight route. According to the layout of the scenic spot and the image content to be collected, plan the flight route on the UAV's flight control software. The flight route should cover the main area of ​​the scenic spot and try to obtain panoramic images and images overlooking the scenic spot from different heights and angles; S104 Network resource supplement collection, Search for publicly available image resources related to tourist attractions on the Internet as a supplement. The Internet resources come from the official websites of the attractions, travel forums, and social media platforms.

8. A tourist attraction recommendation system using virtual reality technology according to claim 7, characterized in that During the preparation of the S102 acquisition device, the relationship between the flight altitude and the shooting range Assume that the camera lens on the drone has a viewing angle of θ, and a horizontal viewing angle is adopted. If the flight altitude is h, then in an ideal situation, the radius of the ground range that the drone can shoot in the horizontal direction can be calculated by the following formula For example, if the horizontal viewing angle θ of the drone lens is 60° and the flight altitude h is 100 meters, then the radius r of the ground range that can be shot in the horizontal direction is That is, in the horizontal direction, a circular ground range with a radius of 57.7 meters centered directly below the drone can be shot.

9. The tourism attraction recommendation system using virtual reality technology according to claim 2, characterized in that, In the image modeling unit, a three-dimensional scene model of the tourist attraction is constructed using three-dimensional modeling technology based on the scenic image information, including the following steps S201 Image acquisition Collect image materials of tourist attractions from multiple data sources to ensure that the collected images can cover different areas, different angles, and multiple situations in different seasons of the attraction, so as to obtain comprehensive visual information; Conduct preliminary screening and sorting of the collected images, remove images that are blurred, have a high degree of duplication and cannot be used for modeling, and retain representative and high-quality images as the basic materials for subsequent modeling; S202 Feature extraction For each screened image, use the feature extraction algorithm in computer vision technology to extract the key feature points and their descriptions in the image. These feature points contain the position information of unique visual characteristics in the image and are used for subsequent feature matching and three-dimensional reconstruction; store the feature points and descriptions of all the extracted images to form a feature database for convenient query and use in subsequent steps; S203 Image registration Based on the previously extracted feature points and descriptions, find the corresponding relationships between different images through feature matching algorithms. Through these algorithms, find the same or similar feature points in two or more images taken from different perspectives, so as to determine the relative position and pose relationships between these images. Further optimize the successfully matched image pairs to obtain a set of accurately registered image sequences, which will provide basic data for subsequent three-dimensional reconstruction; S204 Three-dimensional reconstruction According to the accurately registered image sequences, adopt a three-dimensional reconstruction algorithm based on multi-view geometry. For each set of feature points that are successfully matched in different images, use the internal parameters of the camera when these images are taken. The internal parameters include the focal length and the position of the principal point, and the known relative external parameters. The external parameters include the rotation and translation relationships, and calculate the coordinate position of the feature point in the three-dimensional space through the principle of triangulation; Gradually construct the sparse three-dimensional point cloud model of the entire tourist attraction; S205 Point cloud optimization and densification Perform optimization processing on the generated sparse three-dimensional point cloud model. Adopt statistical filtering and radius filtering methods. According to the distribution law and distance relationship of the points in the point cloud, identify and remove those abnormal points that are significantly inconsistent with the surrounding points to improve the quality of the point cloud model, and convert the sparse point cloud model into a dense point cloud model; S206 Mesh generation Based on the optimized dense point cloud model, the point cloud is converted into a 3D model in the form of a triangular mesh or a polygon mesh; during the mesh generation process, according to the density and shape characteristics of the point cloud, the mesh cells are reasonably divided to ensure that the generated mesh model can accurately reflect the shape of the scenic spot and has a good topological structure, facilitating subsequent rendering and interaction operations; S207 Texture mapping, Select a suitable image from the initially collected high-quality image materials as the texture map, and accurately map the selected image onto the surface of the 3D model according to the mesh structure of the 3D model and the corresponding real-scene positions of each part; finally, a complete 3D scene model of the tourist scenic spot is constructed.

10. A tourist attraction recommendation system using virtual reality technology according to claim 9, characterized in that, In the image acquisition of S201, the acquired images are preliminarily screened and sorted through a preset threshold T V Specifically, for a grayscale image I(x, y), if it is a color image, it needs to be converted to a grayscale image first I(x,y) = 0.299R(x,y) + 0.587G(x,y) + 0.114B(x,y); where R(x,y), G(x,y), and B(x,y) are the red, green, and blue channel values of the color image at this point, respectively; First, calculate the gradient magnitude G(x,y) of the image, and use the Sobel operator to calculate the gradients in the horizontal and vertical directions: Horizontal direction gradient G x (x, y): G x (x,y) = I(x + 1,y) - I(x - 1,y); Vertical direction gradient G y (x, y): G y (x, y) = I(x, y + 1) - I(x, y - 1); Then the gradient magnitude G(x,y) is: Then calculate the variance V of the gradient magnitude of the entire image: where M and N are the number of rows and columns of the image respectively, is the average value of the image gradient magnitude, and the calculation formula is: When the variance V of the magnitude of the entire image gradient is less than the preset threshold T V , it is considered that the image is blurred and is removed.

Images