Visual preference-based garden scenic spot recommendation method, system and equipment, storage medium and head-mounted display equipment

Abstract

The invention provides a visual preference-based garden scenic spot recommendation method, system and device, a storage medium and a head-mounted display device, and relates to the technical field of communication. The method comprises the following steps: obtaining garden landscape data composed of landscape data of a plurality of garden scenic spots and visited data of visited garden scenic spots of tourists; determining a scenic spot label value based on the garden landscape data and the visited data; determining tourist interest values of the tourists interested in different garden scenic spots based on current tour data of the tourists in visiting the garden scenic spots and the visited data; and based on the scenic spot label value and the tourist interest value, determining a scenic spot recommendation set for recommending all garden scenic spots to the tourist. According to the embodiment provided by the invention, interested scenic spots are accurately recommended for tourists, and the sightseeing experience of the tourists is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of communication technologies, and in particular, to a garden scenic spot recommendation method, system, device, storage medium, and head-mounted device based on visual preferences. Background Art

[0002] In related technologies, collaborative filtering, content-based recommendation, and hybrid recommendation are the three main current recommendation techniques. Collaborative filtering makes recommendations by analyzing the behavioral similarities between users or the similarities between items, while content-based recommendation relies on the feature information of items and realizes recommendations by matching the user's profile and item features. Hybrid recommendation combines the advantages of the former two to improve the accuracy and coverage of recommendations. In recent years, the application of deep learning technologies in recommendation systems has also become increasingly widespread, such as using neural networks to extract complex feature representations, and sequence model-based recommendation systems, such as recurrent neural networks (RNNs) and Transformer architectures, which can process time series data and capture users' dynamic preferences.

[0003] Research on visual preferences involves multiple fields such as computer vision, psychology, and cognitive science. In terms of computer vision, image processing and feature extraction techniques are used to analyze various aspects of visual content, such as color, shape, texture, and spatial layout. Deep learning models, especially convolutional neural networks, have achieved remarkable success in image recognition and classification tasks, providing powerful tools for visual preference analysis. In the fields of psychology and cognitive science, researchers explore the mechanisms of human visual attention and the formation reasons of visual preferences through experimental methods such as eye movement tracking and electroencephalogram (EEG).

[0004] When a user is traveling and wants to visit a scenic spot park, there are usually multiple scenic spots in the park for tourists to visit. Usually, a park tour map is distributed to tourists, and there is a simple introduction to each sub-scenic spot on the tour map, so that tourists can go to the scenic spots they are interested in to play. However, the park area is usually large, the scenic spot introductions on the park tour map are not comprehensive, and the scenic spots that tourists are interested in are often not concentrated in one area. Therefore, when tourists blindly go to the scenic spots they are interested in based on the simple scenic spot introduction information, they may find that they are not the scenic spots they are interested in, and the scenic spots that they think are not very interesting based on the simple scenic spot introduction information may actually be the scenic spots that tourists are interested in due to the incomplete introduction. Based on the above example scenarios, in the prior art, it is impossible to accurately recommend the scenic spots that tourists are interested in, especially for the recommendation of garden scenic spots, which usually gives tourists a low sense of experience. Summary of the Invention

[0005] The present invention provides a method, system, device, storage medium and head-mounted device for recommending garden scenic spots based on visual preference, which can accurately recommend scenic spots that tourists are interested in and significantly improve the tourists' tour experience.

[0006] In a first aspect, the present invention provides a method for recommending garden scenic spots based on visual preference, including the following steps: Obtain garden landscape data composed of landscape data of multiple garden scenic spots and visited data of the garden scenic spots that the tourist has visited; Based on the garden landscape data and the visited data, determine the scenic spot label value; Based on the current visit data of the tourist visiting the garden scenic spots and the visited data, determine the tourist interest value of the tourist in different garden scenic spots; Based on the scenic spot label value and the tourist interest value, determine a scenic spot recommendation set for recommending all garden scenic spots to the tourist.

[0007] Preferably, according to a method for recommending garden scenic spots based on visual preference provided by the present invention, the determining the scenic spot label value based on the garden landscape data and the visited data includes: Preprocess the garden landscape data to determine observable garden features; Analyze and process the visited data based on the grid analysis method to determine reachable garden features; Perform weighted calculation processing based on the observable garden features and the reachable garden features to determine the scenic spot label value.

[0008] Preferably, according to a method for recommending garden scenic spots based on visual preference provided by the present invention, the preprocessing the garden landscape data to determine observable garden features includes: Analyze and process the garden landscape data to determine the tourist's preference for the garden landscape during the tour; Classify the landscape data of multiple garden scenic spots based on the preference for the garden landscape during the tour to obtain landscape data with different labels; Use a preset semantic recognition model to identify and process the landscape data with different labels to obtain a landscape recognition result, and use different colors and the landscape recognition result to label the landscape data with different labels to obtain the observable garden features.

[0009] Preferably, according to a method for recommending garden scenic spots based on visual preference provided by the present invention, the visited data at least includes: the displacement speed of the user visiting the scenic spot; The analyzing and processing the visited data based on the grid analysis method to determine reachable garden features includes: Based on the grid analysis method, the garden landscape data is divided into grids to obtain the landscape data of garden scenic spots in different grids; Based on the road data between each garden scenic spot and the displacement speed of the user visiting the scenic spot, calculate the effective visiting distance of the tourist from each garden scenic spot to the target garden scenic spot; Count the number of grid intersections within the effective visiting distance to determine the reachable garden features.

[0010] Preferably, according to a method for recommending garden scenic spots based on visual preference provided by the present invention, the current visit data at least includes: the current stay duration of the current scenic spot at the current visit location, and the current fixation duration of the current scenery at the current visit; The visited data at least includes: the visited stay duration of the visited scenic spot at the visited scenic spot location, and the visited fixation duration of the visited scenery; Based on the current visit data and the visited data of the tourist visiting the garden scenic spot, determine the tourist interest value of the tourist in different garden scenic spots, including: Based on the current stay duration of the current scenic spot and the visited stay duration of the visited scenic spot, determine the scenic spot interest score of the tourist's interest in the garden scenic spot; Based on the current fixation duration of the current scenery and the visited fixation duration of the visited scenery, determine the scenery interest score of the tourist's interest in the scenery of the garden scenic spot; Based on the scenic spot interest score and the scenery interest score, determine the tourist interest value of the tourist in different garden scenic spots.

[0011] Preferably, according to a method for recommending garden scenic spots based on visual preference provided by the present invention, based on the scenic spot label value and the tourist interest value, determine the scenic spot recommendation set for recommending all garden scenic spots to the tourist, including: Based on the scenic spot label value and the tourist interest value, calculate the interest matching degree between the tourist and each garden scenic spot; Sort the interest matching degrees between the tourist and each garden scenic spot to obtain an interest sorting result; Based on the interest sorting result, perform a sorting process on all garden scenic spots to determine the scenic spot recommendation set for recommending all garden scenic spots to the tourist.

[0012] In a second aspect, the present invention also provides a system for recommending garden scenic spots based on visual preference, including: An acquisition data module for acquiring garden landscape data composed of landscape data of multiple garden scenic spots and visited data of the tourist's visited garden scenic spots; A determining scenic spot label value module for determining the scenic spot label value based on the garden landscape data and the visited data; A module for determining tourist interest values, which is used to determine the tourist interest values of tourists in different garden scenic spots based on the current tour data of the tourists visiting the garden scenic spots and the already visited data. A module for determining a scenic spot recommendation set, which is used to determine a scenic spot recommendation set for all garden scenic spots recommended to the tourist based on the scenic spot label values and the tourist interest values.

[0013] In a third aspect, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the method for recommending garden scenic spots based on visual preference as described in any one of the above.

[0014] In a fourth aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the method for recommending garden scenic spots based on visual preference as described in any one of the above.

[0015] In a fifth aspect, the present invention also provides a head-mounted display device, including a camera component and a positioning component. Among them, the camera component is used to determine the gaze behavior of the user, the positioning component is used to determine the tour location of the user, and it also includes the system for recommending garden scenic spots based on visual preference described in the second aspect.

[0016] In a sixth aspect, the present invention also provides a computer program product, including a computer program. When the computer program is executed by a processor, it implements the method for recommending garden scenic spots based on visual preference as described in any one of the above.

[0017] A method, system, device, storage medium, and head-mounted display device for recommending garden scenic spots based on visual preference provided by the present invention. By obtaining garden landscape data composed of landscape data of multiple garden scenic spots and already visited data of the gardens that tourists have visited; determining scenic spot label values based on the garden landscape data and the already visited data; determining the tourist interest values of tourists in different garden scenic spots based on the current tour data of the tourists visiting the garden scenic spots and the already visited data; and determining a scenic spot recommendation set for all garden scenic spots recommended to the tourist based on the scenic spot label values and the tourist interest values. It can accurately recommend scenic spots that tourists are interested in and significantly improve the tourist experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0019] Figure 1 It is one of the flow schematic diagrams of a garden scenic spot recommendation method based on visual preference provided by the present invention.

[0020] Figure 2 It is a schematic diagram of the analysis of garden landscape data provided by the present invention.

[0021] Figure 3 It is a schematic diagram of color annotation for garden scenic spots provided by the present invention.

[0022] Figure 4 It is a flow schematic diagram of determining the scenic spot label value provided by the present invention.

[0023] Figure 5 It is a flow schematic diagram of determining the tourist interest value provided by the present invention.

[0024] Figure 6 It is a flow schematic diagram of determining the scenic spot recommendation set provided by the present invention.

[0025] Figure 7 It is a schematic diagram of the user interaction interface provided by the present invention.

[0026] Figure 8 It is a structural schematic diagram of a garden scenic spot recommendation system based on visual preference provided by the present invention.

[0027] Figure 9 It is a structural schematic diagram of the electronic device provided by the present invention. Detailed implementation manners

[0028] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below with reference to the accompanying drawings in the present invention. Apparently, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments in the present invention belong to the scope of protection of the present invention.

[0029] In the related art, there are at least the following technical problems: Although recommendation algorithms and visual preference technologies have made remarkable progress in providing personalized experiences for users, there are still some deficiencies and limitations in their research and applications. First, recommendation algorithms generally rely on users' historical behavior data, which may not fully reflect users' true preferences. Especially in scenarios where user behavior data is sparse or for new users, the accuracy of the algorithms will be affected. Second, visual preference analysis usually requires complex image processing and pattern recognition technologies, which may be affected by factors such as environmental changes, lighting conditions, and occlusions in actual applications, resulting in a decrease in recognition accuracy.

[0030] On the other hand, the limitations of visual preference analysis technology are that it usually requires a large amount of labeled data to train the model, and the acquisition cost of this data is high and time-consuming. In addition, visual preference is highly subjective, and different users may have very different visual perceptions of the same scenic spot, which makes it difficult to build a generally applicable visual preference model. At the same time, existing visual preference analysis technologies often ignore the influence of non-visual factors such as cultural background and personal experience, which play important roles in users' scenic spot selection.

[0031] The following combines Figures 1-9 to describe a garden scenic spot recommendation method, system, device, storage medium and head-mounted device based on visual preference of the present invention, so as to accurately recommend scenic spots of interest to tourists and significantly improve the tourist experience.

[0032] Figure 1 is one of the schematic flowcharts of a garden scenic spot recommendation method based on visual preference provided by the present invention. As Figure 1 shown, the method may include but is not limited to steps S100 to S400: S100, obtaining garden landscape data composed of landscape data of multiple garden scenic spots and visited data of the garden scenic spots visited by tourists; S200, determining scenic spot label values based on the garden landscape data and the visited data; S300, determining tourist interest values of tourists in different garden scenic spots based on the current visited data of the tourists visiting the garden scenic spots and the visited data; S400, determining a scenic spot recommendation set for recommending all garden scenic spots to the tourists based on the scenic spot label values and the tourist interest values.

[0033] In step S100 of some embodiments, garden landscape data composed of landscape data of multiple garden scenic spots and visited data of the garden scenic spots visited by tourists are obtained.

[0034] In the embodiments of the present invention, the source of the garden landscape data composed of landscape data of multiple garden scenic spots can utilize web crawler technology to crawl garden-related content shared by users on social media and travel websites. This content may include photos taken by tourists, travelogues, reviews, etc., from which information about the garden landscape can be extracted.

[0035] In addition, professional garden photography accounts, travel bloggers, etc. can also be followed, and high-quality garden landscape pictures and detailed introductions can be obtained, which can be used as supplementary data sources.

[0036] The visited data of the scenic spots in the garden that the tourists have visited can be the data of some of the scenic spots that the tourists have visited when touring the garden scenic area, or the data of some of the scenic spots that other tourists have only visited when touring the garden scenic area. Usually, it is the data of the scenic spots that the tourists are interested in, and no specific limitation is made here.

[0037] In some embodiments, the obtained data can also be classified and sorted according to the preset landscape type tags (such as water scenery, flower and wood vegetation, indoor furnishings, rockeries). Similar landscape elements are grouped into one category, and corresponding database fields are established. For example, for plant landscapes, information such as plant names, families and genera, ornamental characteristics (such as flowering periods, leaf colors, etc.), and planting locations are recorded; for water scenery, information such as water body types (lakes, streams, fountains, etc.), areas, water qualities, and surrounding landscapes are recorded.

[0038] Clean the data to remove duplicate, incorrect, or invalid data. For example, check whether the spelling of the plant names is correct and whether the coordinate data is reasonable.

[0039] In step S200 of some embodiments, based on the garden landscape data and the visited data, determine the scenic spot label value.

[0040] It can be understood that after the steps of step S100 are executed, the specific execution steps can be to preprocess the garden landscape data to determine the observable garden features; Analyze and process the visited data based on the grid analysis method to determine the reachable garden features; Perform weighted calculation processing based on the observable garden features and the reachable garden features to determine the scenic spot label value.

[0041] Furthermore, the preprocessing of the garden landscape data to determine the observable garden features includes: Analyze and process the garden landscape data to determine the tourists' preferences for the garden landscapes they visit; Classify the landscape data of multiple garden scenic spots based on the preferences for the garden landscapes they visit to obtain scenic data with different tags; Use a preset semantic recognition model to identify the scenic data with different tags to obtain a scenic recognition result, and use different colors and the scenic recognition result to label the scenic data with different tags to obtain the observable garden features.

[0042] It can be understood that, as shown in combination with Figure 2 Analyze and process the garden landscape data through a computer program or a computer device to determine the tourists' preferences for the garden landscapes they visit.

[0043] Identify and analyze the scenic spot elements and scenery elements in the garden landscape data, and count the appearance frequencies of scenic spots and scenery in the panoramic views taken by tourists at different viewpoints to determine the garden landscape preferences of tourists.

[0044] In some embodiments of the present invention, a large number of tourists' tour history records in the garden are collected, including information such as the scenic spot visit order, stay time, and tour frequency.

[0045] Perform data mining on these data, such as using the association rule mining algorithm, to find the association preferences between different scenic spots of tourists. For example, it is found that most tourists tend to go to the water feature area after visiting the flower exhibition area.

[0046] Analyze the changes in tourists' tour behaviors in different seasons and time periods to determine seasonal and temporal tour preferences. For example, in spring, tourists may be more inclined to visit scenic spots with flower exhibitions.

[0047] Furthermore, according to the determined types of tourists' tour preferences, classify the landscape data of garden scenic spots. For example, classify the landscapes into two major categories: natural landscapes (such as mountains, waters, and plants) and cultural landscapes (such as buildings and sculptures).

[0048] Further subdivide each category of landscape. For example, divide the plant landscape into flower landscapes, tree landscapes, etc., and divide the natural landscape into mountain landscapes, water feature landscapes, etc. according to the landform.

[0049] Even further, garden experts and relevant professionals can be invited to review and adjust the classified landscape data to ensure the accuracy and rationality of the classification.

[0050] In some embodiments, the observable garden features obtained by identifying and annotating the scenery data using a preset semantic recognition model can be: The semantic recognition model is obtained by training on a scene data set. The semantic recognition model can be a text classification model or an image recognition model based on machine learning. For example, for the landscape data described in text, use the support vector machine (SVM) model; for image data, use the convolutional neural network (CNN) model.

[0051] The scene dataset can be ADE20K, which is a large-scale scene parsing dataset. This dataset was created by the Computer Science and Artificial Intelligence Laboratory (CSAIL) at the Massachusetts Institute of Technology (MIT). It contains a wide range of scene and object categories, covering various scenes such as indoor, outdoor, natural, urban, and domestic. The ADE20K dataset contains up to 150 categories of objects, such as buildings, furniture, animals, roads, etc., covering the common object categories in scene parsing tasks. Each image not only has pixel-level semantic labels but also can include instance-level segmentation, making it very suitable for a variety of computer vision tasks.

[0052] Input the classified scenery data into the semantic recognition model to identify the features of the scenery. For example, for the data of flower landscapes, the model can identify features such as the type, color, and flowering period of the flowers.

[0053] According to the scenery recognition results, use different colors or other visualization markers to label the scenery data. For example, label popular landscapes in red and label the water parts in natural landscapes in blue.

[0054] As Figure 3 shown, a schematic diagram of color-labeling garden scenic spots. For example, the blue square is a water landscape, the green square is a flower and tree landscape, the yellow is a rockery landscape, and the orange is an indoor furnishings landscape. The dashed circles in the figure are isochrones drawn for each garden scenic spot, and the solid lines in the figure represent the roads between the garden scenic spots.

[0055] For each scenic spot, intercept panoramic views in different directions. Using the dataset, semantic recognition can be performed on the water, stones, and plants in the pictures and then divided into color blocks of different colors. By the size of the pixels of the color blocks, the landscape label ratio of each picture can be obtained, and then the observable garden features of the scenic spot can be characterized.

[0056] In some embodiments of the present invention, the visited data at least includes: the displacement speed of the user visiting the scenic spot; Analyze and process the visited data based on the grid analysis method to determine the reachable garden features, including: Perform grid division processing on the garden landscape data based on the grid analysis method to obtain the landscape data of garden scenic spots in different grids; Based on the road data between each garden scenic spot and the displacement speed of the user visiting the scenic spot, calculate the effective visiting distance of the tourist from each garden scenic spot to the target garden scenic spot; Count the number of grid intersections within the effective visiting distance to determine the reachable garden features.

[0057] It is understandable that the garden landscape data is gridded and the entire garden space is divided into grids according to certain rules. The size of the grid can be adjusted according to the scale of the garden and the analysis accuracy requirements. For example, for a large garden, the grid can be divided into squares with a side length of 3 meters.

[0058] The landscape data of the garden attractions are mapped to the corresponding grids, and each grid contains the relevant information of the attractions in its area, such as the attraction type, coordinates, etc.

[0059] The role of grid division is to discretize the continuous garden space into small units, making complex spatial data processing simpler and more efficient. For example, it is easier to calculate the activities of tourists in each grid. It provides a basic framework for the subsequent calculation of tourists' effective tour distance and the number of grid intersections, which helps to accurately analyze the movement and tour patterns of tourists in the garden.

[0060] Collect road data within the garden, including road length, width, direction and other information, and integrate the road data with grid data.

[0061] Obtain the displacement speed data of users when visiting scenic spots, which can be collected through mobile phone positioning or MR head display equipment. For example, the MR head display device includes a camera component and a positioning component, wherein the camera component is used to determine the user's gaze behavior, and the positioning component is used to determine the user's visiting position. The movement speed of tourists is determined by analyzing the changes in their visiting positions.

[0062] According to the grid where the tourist's starting and target attractions are located, combined with road data and displacement speed, a path planning algorithm (such as the Dijkstra algorithm) is used to calculate the effective tour distance of tourists from each garden attraction to the target garden attraction. The effective tour distance takes into account the actual conditions of the road and the walking speed of tourists, and can more truly reflect the tourists' tour behavior.

[0063] In the process of calculating the effective tour distance of tourists, the grid sets that tourists pass through are recorded.

[0064] Count the number of grid intersections that tourists pass through at different attractions. For example, for attraction A, construct a 3-minute path isochronous circle with attraction A as the center, and calculate the different values of the three labels in the grids covered by the circle (number of grids), that is, count the number of grid intersections between the isochronous circle and the grid where the label is located.

[0065] The accessibility between scenic spots is determined based on the number of grid intersections. If the number of grid intersections is large, it means that there are many scenic spots that can be reached from the scenery with scenic spot A as the center. Otherwise, the accessibility is low.

[0066] In some embodiments of the present invention, a 3m×3m grid is constructed for the planar layout of the entire garden. According to the existing surveying and mapping data, each grid is assigned three landscape tags: water, stone, and plant (if the scenery is distributed in the grid, it is assigned a value, which is represented by coloring the grid in the figure. Plants are divided into low shrubs and tall trees and assigned different values). Then, all the actual passable roads in the garden are drawn on the plane. Taking each scenic spot as the center, a 3min path isochrone circle is constructed, and the different values of the three tags in the grids covered by the circle are calculated (counting the grids), so as to characterize the "reachable" feature of the landscape of this scenic spot.

[0067] In some embodiments of the present invention, in combination with Figure 4 As shown, the steps of determining the scenic spot tag value through weighted calculation and processing based on the observable garden feature and the reachable garden feature may specifically be: First, determine the weight coefficients of the observable garden feature and the reachable garden feature. The determination of the weight coefficients can refer to expert opinions, tourist surveys, or data analysis results. For example, if it is found through a questionnaire survey that tourists think the aesthetics of the landscape (observable feature) is more important than convenience (reachable feature), then a higher weight can be assigned to the aesthetics. The weight coefficients are not specifically limited here and can be flexibly selected according to the actual needs of tourists.

[0068] Quantitatively score the observable garden feature and the reachable garden feature of each scenic spot. For example, for the observable feature, scores can be given according to aspects such as the uniqueness and richness of the landscape; for the reachable feature, scores can be given according to factors such as the effective tour distance and the number of grid intersections calculated previously.

[0069] According to the weight coefficients and the quantitative scores, use the weighted calculation method to obtain the tag value of each scenic spot. For example, scenic spot tag value = observable feature score × observable feature weight + reachable feature score × reachable feature weight.

[0070] Considering the factors of both observability and reachability comprehensively, comprehensively evaluate the value of each scenic spot, and provide a unified standard for the classification, recommendation, and ranking of scenic spots. For example, a scenic spot with both high aesthetics and high reachability will obtain a higher tag value and can be highly recommended to tourists.

[0071] In step S300 of some embodiments, based on the current tour data and the already visited data of tourists visiting the garden scenic spots, determine the tourist interest value of tourists' interest in different garden scenic spots.

[0072] First, it should be noted that the current tour data at least includes: the current scenic spot stay duration at the current tour location and the current scenery fixation duration of the currently visited scenery.

[0073] The visited data at least includes: the residence duration of the visited scenic spots at the positions of the visited scenic spots, and the fixation duration of the visited scenery for the visited scenery.

[0074] Using the eye tracking technology of the MR headset device (such as a head-mounted eye tracker, a glasses-type eye tracker, etc.), monitor and record the gaze behavior of tourists at appropriate positions in the garden scenic spots (such as near the landscape focus, exhibition items, etc.). These devices can accurately record the start time, end time, and fixation duration of tourists' gazes at different scenery.

[0075] It can be understood that after performing the steps of step S200, the specific execution steps may include: Based on the current scenic spot residence duration and the visited scenic spot residence duration, determine the scenic spot interest score of the tourists' interest in the garden scenic spot; Based on the current scenery fixation duration and the visited scenery fixation duration, determine the scenery interest score of the tourists' interest in the scenery of the garden scenic spot; Based on the scenic spot interest score and the scenery interest score, determine the tourist interest value of the tourists' interest in different garden scenic spots.

[0076] It can be understood that in combination with Figure 5 As shown, a schematic diagram for determining the tourist interest value. Process the current scenic spot residence duration according to a pre-set rule or algorithm. For example, different time intervals can be set to correspond to different scores, such as getting 5 points for a residence duration of more than 30 minutes, 3 points for 15 - 30 minutes, 1 point for 5 - 15 minutes, etc. A more complex functional relationship can also be used to calculate the score to more precisely reflect the relationship between the interest level and the residence duration.

[0077] Similarly, calculate the scores for the visited scenic spot residence durations according to similar rules.

[0078] Furthermore, similar to the position determination, process the current scenery fixation duration and calculate the score according to a pre-set rule or algorithm. For example, getting 5 points for a fixation duration exceeding 10 seconds, 3 points for 5 - 10 seconds, 1 point for 2 - 5 seconds, etc. Different scoring criteria can also be set for different scenery according to factors such as the importance and rarity of the scenery.

[0079] Perform corresponding score calculations for the visited scenery fixation durations as well.

[0080] Fuse the current scene fixation duration score and the fixation duration score of the visited scenes to determine the scene interest score. For example, using the weighted average method, let the weight of the current scene fixation duration be 0.6 and the weight of the visited scene fixation duration be 0.4. Then the scene interest score = 0.6 × current scene fixation duration score + 0.4 × visited scene fixation duration score.

[0081] Furthermore, further analyze and process the scenic spot interest score and the scene interest score. It can be considered to perform a weighted sum of the scenic spot interest score and the scene interest score according to a certain ratio to obtain the comprehensive interest value of the tourist for different garden scenic spots. For example, let the weight of the scenic spot interest score be 0.5 and the weight of the scene interest score be 0.5. Then the interest value of the tourist for a certain garden scenic spot = 0.5 × the scenic spot interest score of this scenic spot + 0.5 × the average scene interest score of the main scenes within this scenic spot.

[0082] The main scenes may include but are not limited to flowers and plants, rockeries, water, and indoor furnishings.

[0083] It is also possible to set different weight allocation schemes for different scenic spots and scenes according to factors such as the type and importance of the scenic spots and scenes. For example, for garden scenic spots mainly composed of natural landscapes, the weight of the scenic spot interest score may be higher; while for scenic spots mainly for cultural exhibitions, the weight of the scene interest score may be higher.

[0084] By comprehensively considering the scenic spot interest score and the scene interest score, a more comprehensive and accurate tourist interest value can be obtained. This interest value not only reflects the overall interest degree of the tourist in the garden scenic spot, but also takes into account the tourist's attention to the specific scenes within the scenic spot, which helps to more accurately grasp the tourist's interest preferences.

[0085] In step S400 of some embodiments, based on the scenic spot label value and the tourist interest value, determine the scenic spot recommendation set for recommending all garden scenic spots to the tourist.

[0086] It can be understood that after executing the steps of step S300, the specific execution steps can be: based on the scenic spot label value and the tourist interest value, calculate the interest matching degree between the tourist and each garden scenic spot; Sort the interest matching degrees between the tourist and each garden scenic spot to obtain the interest sorting result; Based on the interest sorting result, perform a sorting process on all garden scenic spots to determine the scenic spot recommendation set for recommending all garden scenic spots to the tourist.

[0087] Furthermore, in combination with Figure 6As shown, the scenic spot label values calculated through step S200 and the tourist interest values obtained through step S300 are calculated and processed to calculate the interest matching degree between tourists and each garden scenic spot.

[0088] In an embodiment of the present invention, a recommendation algorithm can be used to process the scenic spot label values and the tourist interest values to calculate the interest matching degree between tourists and each garden scenic spot.

[0089] Specifically, the KL divergence can be used to compare the similarity to calculate the similarity measure between tourists and different scenic spots. By comparing the similarity through the KL divergence, the KL divergence is an asymmetric distance metric for measuring the difference between two probability distributions and is often used to evaluate the performance of recommendation systems.

[0090] The tourist interest values and the scenic spot label values are normalized, and the similarity of the two normalized probability distributions is compared through the KL divergence algorithm to determine the matching degree between different scenic spots and the tourist interests.

[0091] In some embodiments of the present invention, if both the scenic spot label values and the tourist interest values are represented in vector form, the cosine similarity between vectors can also be calculated to measure the matching degree. The value range of the cosine similarity is between -1 and 1. The closer the value is to 1, the higher the matching degree. The closer the value is to -1, the lower the matching degree. 0 indicates that there is no obvious similarity between the two.

[0092] A reasonable matching degree calculation method can accurately quantify the relationship between tourists and scenic spots. The vector space model can vectorize high-dimensional data for easy calculation and comparison. Selecting an appropriate calculation method can more accurately reflect the fit between tourist interests and scenic spot attributes according to specific data characteristics and application scenarios.

[0093] Furthermore, according to the selected matching degree calculation results, all garden scenic spots are sorted in descending order of the interest matching degree with tourists. Common sorting algorithms can be used, such as quicksort, mergesort, heapsort, etc. For example, when using the quicksort algorithm, a pivot element is selected, and the scenic spots corresponding to the matching degrees greater than the pivot element are placed on the left, and those less than the pivot element are placed on the right, and then the left and right parts are recursively sorted respectively.

[0094] Furthermore, according to actual needs, flexibly set the criteria for selecting recommended scenic spots. For example, the upper limit of the number of recommended scenic spots can be set, such as recommending the top 5 or top 10 most matching scenic spots to tourists; factors such as the popularity of the scenic spot, tourist flow restrictions, and seasonal adaptability can also be considered. For example, if a certain scenic spot has a high matching degree with the tourist interests but is currently under maintenance or has reached the maximum carrying capacity, it may not be included in the scenic spot recommendation set.

[0095] In some embodiments of the present invention, clearly defining the criteria for recommending scenic spots can ensure the rationality and feasibility of the recommended set. By restricting the number of recommended scenic spots, it is possible to avoid providing tourists with too many choices, which may cause decision-making difficulties. At the same time, considering other factors such as the status of the scenic spots and their carrying capacity can improve the accuracy and practicality of the recommendations.

[0096] In addition, the selection criteria can also be adjusted individually according to different tourist groups (such as family tourists, couple tourists, elderly tourists, etc.) to meet diverse needs.

[0097] In some embodiments, in combination with Figure 7 As shown, according to the sorting results and selection criteria, determine the set of garden scenic spots finally recommended to tourists. Arrange the selected scenic spots in the order of matching degree or according to a certain display logic (such as combining scenic spots adjacent in geographical location) to form a complete recommended list for presentation to tourists. For example, the scenic spot recommendation set may include, but is not limited to, Scenic Spot A, Scenic Spot B, and Scenic Spot C. Display Scenic Spot A, Scenic Spot B, and Scenic Spot C to tourists through the user interface design and receive the selection feedback from users.

[0098] The generated scenic spot recommendation set directly provides valuable visit guidance for tourists. It helps tourists quickly discover the scenic spots they may be interested in and improves the tourist experience and satisfaction in the garden.

[0099] In some embodiments, display the recommended list of the scenic spot recommendation set of the garden scenic spots to tourists through the user interface design and receive the selection feedback from users.

[0100] Specifically, first, integrate map data: collect detailed map data of the garden scenic area, including information such as the precise location coordinates of scenic spots, road directions, and building distributions. These data can be sorted and digitally processed through Geographic Information System (GIS) technology for accurate presentation on the user interface.

[0101] Secondly, draw and update the map: use professional map drawing software or tools to convert the processed map data into a visual map interface. The map should have clear markings, such as the names of scenic spots and road names, for easy identification by tourists. And a regular update mechanism should be established to ensure that the map data is consistent with the actual situation of the scenic area. For example, when new scenic spots are added or road construction occurs in the scenic area, the map should be updated in a timely manner.

[0102] Finally, highlight the recommended scenic spots: According to the previously determined scenic spot recommendation set, highlight these recommended scenic spots on the map with special icons, colors, or animation effects. For example, a golden flash icon can be used to represent highly matched recommended scenic spots, and a green icon can be used to represent moderately matched recommended scenic spots, so that tourists can notice the locations of the recommended scenic spots at a glance.

[0103] Furthermore, parameter setting interface design: Create a dedicated personalized setting area in the user interface, such as called "Recommendation Preferences". In this area, provide a series of parameter options related to the recommendation algorithm for users to choose. These parameters can include the weights of scenic spot types (such as the weights of natural landscapes, cultural and historical sites, entertainment facilities, etc.), the preference for tour time (such as preferring short tours or long in-depth tours), the preference for environmental atmosphere (such as quiet, lively, etc.).

[0104] Even further, users can adjust the values or options of each parameter through methods such as sliders, drop-down menus, etc. The system should provide real-time feedback on the parameter settings adjusted by the user. For example, when the user increases the weight of natural landscape types, the natural landscape scenic spots in the recommendation list will be ranked correspondingly higher. And users can save their parameter settings in their personal accounts so that the same personalized settings are still maintained when using the application next time.

[0105] In this way, it is possible to recommend scenic spots to meet the personalized needs of different tourists. Different tourists have different interests and preferences, and the personalized setting options allow users to customize the recommendation results according to their own preferences. For example, tourists who like history and culture can increase the weight of cultural and historical sites, making the recommendation list more in line with personal interests, thereby improving the satisfaction of tourists with the recommendations.

[0106] Even further, set multiple feedback collection channels in the user interface. For example, set "Like" and "Dislike" buttons below the introduction page of each recommended scenic spot, and users can click the corresponding buttons to record their initial attitudes towards the scenic spot. At the same time, a text box can also be provided for users to fill in specific evaluations and opinions about the scenic spot, such as the advantages and disadvantages of the scenic spot, the feelings of the tour experience, etc. In addition, after a tourist completes a tour, collect more comprehensive feedback information through a pop-up questionnaire, including the satisfaction with the entire recommendation list, whether the tour expectations are met, etc.

[0107] The user feedback data collected is stored in the background database. The system regularly processes and analyzes these data. For example, count the number of "Likes" and "Dislikes" for each scenic spot, and analyze the keywords and themes mentioned in the user evaluations. Through data analysis, it can be found which scenic spots are more popular among tourists, which scenic spots need improvement, and the opinions and suggestions of users on the recommendation algorithm. Then, adjust the parameters of the recommendation algorithm and optimize the scenic spot recommendation set according to these analysis results to improve the quality of future recommendations.

[0108] Combined with Figure 7 As shown, in some embodiments of the present invention, the method for recommending garden scenic spots based on visual preferences may further include but is not limited to the following steps: a) Provide a mixed reality experience for tourists through MR technology equipment; b) Use an MR headset device to monitor the visual focus points of tourists during the tour; c) Analyze the visual focus points and stay time of tourists through data processing; d) Generate a personalized scenic spot recommendation list according to the visual preferences of tourists through a recommendation algorithm; e) Display the recommendation list to tourists through the user interface design and receive the selection feedback of users. f) Collect the selection and evaluation of users on the recommended scenic spots to optimize future recommendation results.

[0109] In some embodiments of the present invention, to detect the effectiveness of the recommendation strategy, the core steps are transformed into an experiment. By watching panoramic pictures in a VR laboratory, combined with physiological index monitoring and subjective scale filling, its effectiveness is verified. Through a control experiment of multiple comparison algorithms, the most suitable algorithm for users may be selected through satisfaction. The experimental data shows that the scenic spots recommended by the system highly coincide with the locations where tourists actually stay for a long time, indicating that the system can accurately predict and guide tourists to the locations they may be interested in, thereby improving the satisfaction of tourists. Among them, physiological index monitoring such as wearing an electroencephalogram device or a heart rate monitoring device, and the subjective scale is to let tourists fill in their subjective feelings (whether they are interested) about the panoramic pictures of different control groups they see. The two cooperate with each other to verify the satisfaction degree of tourists with the recommendation results.

[0110] In some embodiments of the present invention, the present invention also provides a headset device, including a camera component and a positioning component. Among them, the camera component is used to determine the gaze behavior of the user, the positioning component is used to determine the tour location of the user, and it also includes the garden scenic spot recommendation system based on visual preference described below.

[0111] A garden scenic spot recommendation method, system, device, storage medium and headset device provided by the present invention, by obtaining garden landscape data composed of landscape data of multiple garden scenic spots and the visited data of the garden scenic spots that tourists have visited; based on the garden landscape data and the visited data, determine the scenic spot label value; based on the current visited data of tourists visiting the garden scenic spots and the visited data, determine the tourist interest value of tourists in different garden scenic spots; based on the scenic spot label value and the tourist interest value, determine the scenic spot recommendation set for recommending all garden scenic spots to the tourists. It realizes accurately recommending scenic spots that tourists are interested in, and significantly improves the tourist experience.

[0112] The garden scenic spot recommendation system based on visual preference provided by the present invention will be described below. The garden scenic spot recommendation system based on visual preference described below can be mutually corresponding and referred to the garden scenic spot recommendation method described above.

[0113] As Figure 8The following is a schematic structural diagram of a garden scenic spot recommendation system based on visual preference provided by the present invention. A garden scenic spot recommendation system based on visual preference includes the following modules: A data acquisition module 810, configured to acquire garden landscape data composed of landscape data of multiple garden scenic spots and visited data of the garden scenic spots visited by tourists; A scenic spot label value determination module 820, configured to determine a scenic spot label value based on the garden landscape data and the visited data; A tourist interest value determination module 830, configured to determine tourist interest values of tourists in different garden scenic spots based on current visit data of tourists visiting the garden scenic spots and the visited data; A scenic spot recommendation set determination module 840, configured to determine a scenic spot recommendation set for recommending all garden scenic spots to the tourists based on the scenic spot label value and the tourist interest value.

[0114] Preferably, a garden scenic spot recommendation system based on visual preference provided by an embodiment of the present invention is further specifically configured to preprocess the garden landscape data to determine observable garden features; Analyze and process the visited data based on grid analysis to determine reachable garden features; Perform weighted calculation processing based on the observable garden features and the reachable garden features to determine the scenic spot label value.

[0115] Preferably, a garden scenic spot recommendation system based on visual preference provided by an embodiment of the present invention is further specifically configured to analyze and process the garden landscape data to determine the tourist's preference for the garden landscape during the visit; Classify the landscape data of multiple garden scenic spots based on the preference for the garden landscape during the visit to obtain scenic data with different labels; Use a preset semantic recognition model to recognize the scenic data with different labels to obtain a scenic recognition result, and label the scenic data with different labels using different colors and the scenic recognition result to obtain the observable garden features.

[0116] Preferably, a garden scenic spot recommendation system based on visual preference provided by an embodiment of the present invention is further specifically configured to the visited data at least includes: the displacement speed of the user visiting the scenic spot; Perform grid division processing on the garden landscape data based on the grid analysis method to obtain the landscape data of the garden scenic spots in different grids; Calculate the effective visit distance of the tourist from each garden scenic spot to the target garden scenic spot based on the road data between each garden scenic spot and the displacement speed of the user visiting the scenic spot; Count the number of grid intersections within the effective tour distance to determine the reachable garden features.

[0117] Preferably, a garden scenic spot recommendation system based on visual preference provided by an embodiment of the present invention is specifically further configured such that the current tour data at least includes: the current scenic spot stay duration at the current tour location and the current scenic view fixation duration of the current tour scenery; The already visited data at least includes: the already visited scenic spot stay duration at the already visited scenic spot location and the already visited scenic view fixation duration of the already visited scenery; Based on the current scenic spot stay duration and the already visited scenic spot stay duration, determine the scenic spot interest score that the tourist is interested in the garden scenic spot; Based on the current scenic view fixation duration and the already visited scenic view fixation duration, determine the scenic view interest score that the tourist is interested in the scenery of the garden scenic spot; Based on the scenic spot interest score and the scenic view interest score, determine the tourist interest value that the tourist is interested in different garden scenic spots.

[0118] Preferably, a garden scenic spot recommendation system based on visual preference provided by an embodiment of the present invention is specifically further configured to calculate the interest matching degree between the tourist and each garden scenic spot based on the scenic spot label value and the tourist interest value; Rank the interest matching degrees between the tourist and each garden scenic spot to obtain an interest ranking result; Based on the interest ranking result, perform a ranking process on all garden scenic spots to determine a scenic spot recommendation set for recommending all garden scenic spots to the tourist.

[0119] A garden scenic spot recommendation method, system, device, storage medium, and head-mounted device based on visual preference provided by the present invention obtain garden landscape data composed of landscape data of multiple garden scenic spots and already visited data of the garden scenic spots already visited by the tourist; determine the scenic spot label value based on the garden landscape data and the already visited data; determine the tourist interest value that the tourist is interested in different garden scenic spots based on the current tour data of the tourist visiting the garden scenic spot and the already visited data; determine a scenic spot recommendation set for recommending all garden scenic spots to the tourist based on the scenic spot label value and the tourist interest value. It realizes accurately recommending interesting scenic spots to tourists and significantly improves the tourist's tour experience.

[0120] Figure 9 Illustrates a schematic physical structure diagram of an electronic device, such as Figure 9As shown in the figure, the electronic device may include: a processor 910, a communications interface 920, a memory 930, and a communication bus 940. Among them, the processor 910, the communications interface 920, and the memory 930 communicate with each other through the communication bus 940. The processor 910 can call the logical instructions in the memory 930 to execute the garden scenic spot recommendation method based on visual preferences. The method includes: obtaining garden landscape data composed of landscape data of multiple garden scenic spots and the visited data of the garden scenic spots that the tourist has visited; determining the scenic spot label values based on the garden landscape data and the visited data; determining the tourist interest values of the tourist's interest in different garden scenic spots based on the current visited data of the tourist visiting the garden scenic spots and the visited data; and determining a scenic spot recommendation set for recommending all garden scenic spots to the tourist based on the scenic spot label values and the tourist interest values.

[0121] In addition, when the logical instructions in the above-mentioned memory 930 are implemented in the form of a software functional unit and sold or used as an independent product, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: various media such as a USB flash drive, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk, or an optical disc that can store program codes.

[0122] On the other hand, the present invention also provides a computer program product. The computer program product includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the garden scenic spot recommendation method based on visual preferences provided by the above-mentioned various methods. The method includes: obtaining garden landscape data composed of landscape data of multiple garden scenic spots and the visited data of the garden scenic spots that the tourist has visited; determining the scenic spot label values based on the garden landscape data and the visited data; determining the tourist interest values of the tourist's interest in different garden scenic spots based on the current visited data of the tourist visiting the garden scenic spots and the visited data; and determining a scenic spot recommendation set for recommending all garden scenic spots to the tourist based on the scenic spot label values and the tourist interest values.

[0123] In another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the method for recommending garden scenic spots based on visual preferences provided by the above-mentioned various methods. The method includes: obtaining garden landscape data composed of landscape data of multiple garden scenic spots and the visited data of the garden scenic spots that the tourists have visited; determining the scenic spot label values based on the garden landscape data and the visited data; determining the tourist interest values of the tourists in different garden scenic spots based on the current visit data of the tourists visiting the garden scenic spots and the visited data; and determining a scenic spot recommendation set for recommending all garden scenic spots to the tourists based on the scenic spot label values and the tourist interest values.

[0124] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative work.

[0125] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solution, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0126] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments or equivalently replace some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A garden scenic spot recommendation method based on visual preference, characterized in that, Including: Obtaining garden landscape data composed of landscape data of multiple garden scenic spots and visited data of the garden scenic spots that the tourist has visited; Determining scenic spot label values based on the garden landscape data and the visited data; Determining tourist interest values of the tourist in different garden scenic spots based on the current visit data of the tourist visiting the garden scenic spots and the visited data; Determining a scenic spot recommendation set for recommending all garden scenic spots to the tourist based on the scenic spot label values and the tourist interest values.

2. The method for recommending garden scenic spots based on visual preference according to claim 1, characterized in that The determining the scenic spot label values based on the garden landscape data and the visited data includes: Performing preprocessing on the garden landscape data to determine observable garden features; Performing analysis and processing on the visited data based on the grid analysis method to determine reachable garden features; Performing weighted calculation processing based on the observable garden features and the reachable garden features to determine the scenic spot label values.

3. The method for recommending garden scenic spots based on visual preference according to claim 2, characterized in that The performing preprocessing on the garden landscape data to determine observable garden features includes: Performing analysis and processing on the garden landscape data to determine the tourist's preference for the visited garden landscape; Performing classification processing on the landscape data of multiple garden scenic spots based on the preference for the visited garden landscape to obtain scenic data with different labels; Using a preset semantic recognition model to perform recognition processing on the scenic data with different labels to obtain a scenic recognition result, and using different colors and the scenic recognition result to label the scenic data with different labels to obtain the observable garden features.

4. The method for recommending garden scenic spots based on visual preference according to claim 2, characterized in that The visited data at least includes: the displacement speed of the user visiting the scenic spot; The performing analysis and processing on the visited data based on the grid analysis method to determine reachable garden features includes: Performing grid division processing on the garden landscape data based on the grid analysis method to obtain the landscape data of the garden scenic spots in different grids; Calculating the effective visit distance of the tourist from each garden scenic spot to the target garden scenic spot based on the road data between each garden scenic spot and the displacement speed of the user visiting the scenic spot; Counting the number of grid intersections within the effective visit distance to determine the reachable garden features.

5. The method for recommending garden scenic spots based on visual preference according to any one of claims 1 to 4, characterized in that The current visit data at least includes: the current scenic spot stay duration at the current visit location, the current scenic object fixation duration of the current visited scenery; The visited data at least includes: the visited scenic spot stay duration at the visited scenic spot location, the visited scenic object fixation duration of the visited scenery; The determining tourist interest values of the tourist in different garden scenic spots based on the current visit data of the tourist visiting the garden scenic spots and the visited data includes: Determining the scenic spot interest score of the tourist's interest in the garden scenic spot based on the current scenic spot stay duration and the visited scenic spot stay duration; Based on the current scene fixation duration and the fixation duration of the visited scenes, determine the scene interest score of the scenes that the tourist is interested in for the garden scenic spots; Based on the scenic spot interest score and the scene interest score, determine the tourist interest value of the tourist's interest in different garden scenic spots.

6. The method for recommending garden scenic spots based on visual preference according to any one of claims 1 to 4, characterized in that Based on the scenic spot label value and the tourist interest value, determining a scenic spot recommendation set for recommending all garden scenic spots to the tourist, including: Based on the scenic spot label value and the tourist interest value, calculate the interest matching degree between the tourist and each garden scenic spot; Sort the interest matching degrees between the tourist and each garden scenic spot to obtain an interest sorting result; Based on the interest sorting result, perform a sorting process on all garden scenic spots to determine a scenic spot recommendation set for recommending all garden scenic spots to the tourist.

7. A garden scenic spot recommendation system based on visual preference, characterized in that, Including: An acquisition data module, configured to acquire garden landscape data composed of landscape data of multiple garden scenic spots and visited data of the garden scenic spots visited by the tourist; A determining scenic spot label value module, configured to determine the scenic spot label value based on the garden landscape data and the visited data; A determining tourist interest value module, configured to determine the tourist interest value of the tourist's interest in different garden scenic spots based on the current visit data and the visited data of the tourist visiting the garden scenic spots; A determining scenic spot recommendation set module, configured to determine a scenic spot recommendation set for recommending all garden scenic spots to the tourist based on the scenic spot label value and the tourist interest value.

8. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and running on the processor, characterized in that When the processor executes the program, it implements the method for recommending garden scenic spots based on visual preference according to any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method for recommending garden scenic spots based on visual preference according to any one of claims 1 to 6.

10. A head-mounted display device, comprising a camera assembly and a positioning assembly, wherein, The camera assembly is used to determine the user's fixation behavior, and the positioning assembly is used to determine the user's visit location. Characterized in that it further includes the system for recommending garden scenic spots based on visual preference according to claim 7.

Images