Visual recommendation method and device based on light scene perception of people

Abstract

The invention belongs to the technical field of computers, and provides a sightseeing recommendation method based on light scene perception of people, which comprises the following steps: acquiring a prediction time period corresponding to target sightseeing time and a light climate prediction type, and acquiring travel distance representation between scenic spots in a scenic area; obtaining the light scene experience evaluation of the scenic spot under the prediction time period and the light climate prediction type from the light scene experience relation table; the optimal touring route of the target touring time is determined in combination with the light scene experience evaluation of the scenic spots and the travel distance representation between the scenic spots, and the optimal touring route passes through different scenic spots and takes the entrance of the scenic spots as the starting point and the exit of the scenic spots as the ending point; recommending an optimal touring route to the tourist; the invention further provides a sightseeing recommendation device based on the light scene perception of the person. According to the invention, the optimal touring route can be guided by the perception of people on the scene, the recommendation accuracy can be improved, and the experience satisfaction degree of tourists can be improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a tour recommendation method and device based on human scene perception. Background Art

[0002] In order to improve the tourist experience, the existing metaverse virtual tourism system uses physiological triggering devices to trigger the senses in real time and receive feedback data to virtualize a variety of tourism scenarios and experiences; the existing cultural tourism big data platform centrally dispatches and processes data through the central management module, and at the same time connects the power detection module to achieve a simultaneous detection and display of multiple data. The existing tourism system mainly focuses on creating a rich and multi-level tourism experience and reversely manages the digital universe optimization process and real-time detection of scenic spots through collecting user experience data, dispatching scenic spot information, and tends to be more personal and general in terms of experience. The user's personal subjective (blind) experience is the core premise, lacking the pertinence of human perception, that is, lacking the tourism experience guided by human perception of light and scenery (visual perception), resulting in insufficient satisfaction of tourists, and no targeted management method based on the tourist population has been produced for the management and operation of scenic spots.

[0003] Among them, light scene refers to the landscape composed of light source, scenery, light and shadow and their changes, or refers to the landscape that causes a strong visual impression due to light source, light and shadow and their changes. In the triangular relationship between people, environment and space, light scene emphasizes people's perception under different light environments. Summary of the invention

[0004] The present invention aims to solve the technical problem that the existing tour recommendation methods lack the guidance of people's perception of light and scenery to recommend tour routes, resulting in insufficient tourist experience satisfaction, and provide a tour recommendation method and device based on people's perception of light and scenery.

[0005] In order to achieve the above-mentioned purpose of the present invention, according to the first aspect of the present invention, the present invention provides a tour recommendation method based on human light scene perception, including: obtaining the predicted time period and light climate prediction type corresponding to the target tour time, and obtaining the travel distance representation between scenic spots in the scenic area; obtaining the light scene experience evaluation of the scenic spot under the predicted time period and light climate prediction type from the light scene experience relationship table; combining the light scene experience evaluation of the scenic spot and the travel distance representation between the scenic spots, determining the optimal tour route under the target tour time, the optimal tour route passing through different scenic spots and starting from the entrance of the scenic area and ending at the exit of the scenic area; recommending the optimal tour route to tourists.

[0006] In order to achieve the above-mentioned purpose of the present invention, according to the second aspect of the present invention, the present invention provides a tour recommendation device based on human light scene perception, which is used to implement a tour recommendation method based on human light scene perception described in the first aspect of the present invention, including: a data acquisition module, which obtains the predicted time period and light climate prediction type corresponding to the target tour time, and obtains the travel distance representation between scenic spots in the scenic area; a query module, which obtains the light scene experience evaluation of the scenic spot under the predicted time period and light climate prediction type from the light scene experience relationship table; a module for determining the optimal tour route, which combines the light scene experience evaluation of the scenic spot and the travel distance representation between the scenic spots to determine the optimal tour route under the target tour time, wherein the optimal tour route passes through different scenic spots and starts from the entrance of the scenic area and ends at the exit of the scenic area; and a recommendation module, which recommends the optimal tour route to tourists.

[0007] The beneficial technical effects of the present invention are as follows: the present invention first obtains the predicted time period and light and climate prediction type corresponding to the target sightseeing time, and then queries the light experience evaluation under the predicted time period and light and climate prediction type according to the pre-constructed light experience relationship table, and determines the optimal sightseeing route that can balance the light experience evaluation and the travel distance representation under the target sightseeing time based on the queried light experience evaluation and the distance representation between scenic spots in the scenic area. The obtained optimal sightseeing route can be guided by people's perception of the light scenery (light experience evaluation), rather than the traditional recommendation of sightseeing routes based solely on a wide range of user subjective experience effect satisfaction data analysis. The present invention can improve the recommendation accuracy, thereby improving tourists' experience satisfaction, and facilitates targeted management of scenic area operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 It is a flow chart of a tour recommendation method based on human scene perception in a preferred embodiment of the present invention;

[0009] Figure 2 It is an implementation flow chart of a tour recommendation method based on human scene perception in an application scenario of the present invention;

[0010] Figure 3 It is a principle diagram of measuring indoor illumination information of a scenic spot in a preferred embodiment of the present invention;

[0011] Figure 4 It is a principle diagram of measuring outdoor illumination information of a scenic spot in a preferred embodiment of the present invention;

[0012] Figure 5 It is a schematic diagram of a process for acquiring brightness information of a scenic spot in a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0013] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0014] In the description of the present invention, it is necessary to understand that the terms "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0015] In the description of the present invention, unless otherwise specified and limited, it should be noted that the terms "installed", "connected" and "connected" should be understood in a broad sense. For example, it can be a mechanical connection or an electrical connection, or it can be the internal connection between two components. It can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to the specific circumstances.

[0016] The execution subject of the tour recommendation method based on human light scene perception provided by the present invention includes but is not limited to at least one of the electronic devices such as the server and the terminal that can be configured to execute the method provided in the embodiment of the present application. In other words, the tour recommendation method based on human light scene perception provided by the present invention can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc. The server can be an independent server, or it can be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and big data and artificial intelligence platforms.

[0017] The present invention provides a method for recommending sightseeing based on human scene perception. In a preferred embodiment, Figure 1 and Figure 2 As shown, including:

[0018] Step S1, obtaining the predicted time period corresponding to the target sightseeing time and the light climate prediction type, and obtaining the travel distance representation between scenic spots in the scenic area.

[0019] In this embodiment, the scenic area is preferably, but not limited to, a park, a garden, or an exhibition area. The scenic area includes a plurality of scenic spots, and the travel distance between any two scenic spots is obtained. The travel distance between the two scenic spots is linearly mapped to a dimensionless numerical interval, and the mapping value obtained in the numerical interval is used as the travel distance between the two scenic spots for subsequent calculations. The numerical interval is preferably, but not limited to, [1,2], [1,10], etc. The inventors found by analyzing a large amount of tourist tour data of the scenic area during 9:00-17:00 that dividing the scenic area's tour time into three time periods of morning, noon, and afternoon can basically cover most tourist tour time types. Generally, the morning time period refers to 9:00-12:00, the noon time period refers to 12:00-14:00, and the afternoon time period refers to 14:00-17:00. In an example, assuming that the target tour time is 10:00 on XX / XX / XX, the predicted time period corresponding to the target tour time is morning, and the light climate type corresponding to the morning of XX / XX / XX or 10:00 on XX / XX / XX / XX is obtained from the weather forecast platform as the light climate prediction type. Light climate refers to the sum of the local outdoor illumination conditions and the meteorological factors that affect its changes, and is mainly composed of the average conditions of natural light formed by direct sunlight, diffuse light from the sky, and reflected light from the ground. Light climate types include sunny days with high sun altitude angles, sunny days with low sun altitude angles, fully cloudy days, cloudy to sunny days, cloudy days, rainy days, etc. Different light climate types represent different light intensities, have different light scenes, and tourists' perception and experience are also different.

[0020] Step S2, obtaining the light experience evaluation of the scenic spot under the forecast time period and light climate forecast type from the light experience relationship table.

[0021] A light experience relationship table is pre-built and stored. The light experience relationship table includes the light experience evaluations corresponding to each scenic spot under different light and climate types in different time periods. Each scenic spot has multiple light experience evaluations. A unique light experience evaluation is determined from multiple light experience evaluations based on the predicted time period and light and climate prediction type. The light experience relationship table establishes the connection between the light and scene measured data and people's subjective evaluation.

[0022] Step S3, combining the light experience evaluation of the scenic spot and the travel distance between the scenic spots, determining the optimal tour route under the target tour time, the optimal tour route can balance the light experience evaluation and the travel distance, the optimal tour route passes through different scenic spots and starts at the scenic spot entrance and ends at the scenic spot exit.

[0023] In this embodiment, the scenic area may include multiple scenic area entrances and multiple scenic area exits. This application can preset the scenic area entrance as the main entrance of the scenic area for easy management, and the scenic area exit can be any exit of the scenic area. The optimal tour route passes through different scenic spots, indicating that the scenic spots it contains are not repeated. Generally speaking, on the one hand, the longer the travel distance in the scenic area, the higher the tourist fatigue, and fatigue has a negative impact on the tour experience; on the other hand, being able to visit more interesting attractions helps to improve the satisfaction of the tour experience. To this end, the optimal tour route of this application can balance the light experience evaluation and travel distance representation to maximize the tourist experience satisfaction.

[0024] Step S4, recommending the best tour route to tourists.

[0025] In this embodiment, it is preferred but not limited to recommending the best tour route to tourists through a tourist terminal application (APP) or a display device or broadcasting device at the entrance of the scenic spot.

[0026] In a preferred embodiment, in order to establish an accurate connection between the scene and the subjective evaluation system of tourists, a quantifiable evaluation system is formed for the scene perception, thereby achieving accurate tour route recommendations, such as Figure 2 As shown in the figure, the construction process of the light scene experience relationship table includes:

[0027] Step B1, construct multiple light experience evaluation dimensions. Preferably, four light experience evaluation dimensions are constructed, namely, light comfort, light perception, visual impression and light environment experience. The weights of the four light experience evaluation dimensions are determined, recorded as w1, w2, w3, and w4. Each light experience evaluation dimension is based on the objective environmental conditions of the landscape node. The Likert five-level scale is used to ask questions about the factors that mainly affect a certain light experience evaluation dimension of the scenic spot (4 groups of questions for 4 light experience evaluation dimensions), and the weights of each question in the same light experience evaluation dimension are equal.

[0028] Step B2, obtaining scores of multiple tourists on each light scene experience evaluation dimension of each scenic spot under different light climate types at different time periods.

[0029] The five-level options of the scale are quantified as 5, 4, 3, 2, and 1 respectively. For each scenic spot in each time period and each light climate type, the score of each tourist on each light experience evaluation dimension is obtained as follows: the average score of a group of questions on the light experience evaluation dimension is calculated to obtain the score of the light experience evaluation dimension. The scores of the four light experience evaluation dimensions are recorded as a, b, c, and d respectively.

[0030] In this embodiment, in order to improve the accuracy, the reliability analysis of the Likert five-level scale is performed on the survey results.

[0031] The specific steps are as follows:

[0032] 1) Summarize the answer data of each question in each group of each scene experience evaluation dimension.

[0033] 2) Use SPSSAU to perform reliability analysis and calculate the Cronbach coefficient ε:

[0034]

[0035] Among them, K is the number of questions in each group, S v represents the variance of all tourists’ scores for the vth question in each group of questions, S T 2 It represents the variance of the total scores of K questions for each group of questions by all tourists.

[0036] 3) If the Cronbach's alpha is above 0.7, the questionnaire can be used for subsequent calculation and analysis; otherwise, the scale should be redesigned or the items should be adjusted.

[0037] Step B3, based on the scores of multiple tourists on each light scene experience evaluation dimension of each scenic spot under different light climate types in different time periods, the measured light scene experience evaluation of each scenic spot under different light climate types in different time periods is obtained.

[0038] For each tourist in each time period and each light climate type, the weighted average of the scores of the four light experience evaluation dimensions is calculated according to the weights of the four light experience evaluation dimensions:

[0039] E1=(a*w1+b*w2+c*w3+d*w4) / (w1+w2+w3+w4);

[0040] For each scenic spot, each time period, and each light climate type, the weighted average of the scores of all tourists in the four light experience evaluation dimensions is calculated, and this average is used as the measured light experience evaluation of each scenic spot in each time period and each light climate type.

[0041] The above process is repeated at different time periods under different light climate types (weather) at each scenic spot, thus obtaining the light experience evaluation of each scenic spot in three time periods throughout the day under various light climate types, that is, the measured light experience evaluation of each scenic spot under different conditions.

[0042] Step B4, construct a light experience relationship table using the measured light experience evaluation of each scenic spot in different time periods and different light and climate types. Specifically, the measured light experience evaluation is directly used as the light experience evaluation. The light experience relationship table includes storage trees corresponding to all scenic spots. Each scenic spot corresponds to a four-layer storage tree. The first-layer nodes include only one root node, representing the scenic spot. The second-layer nodes are all connected to the root node of the first layer, and the second-layer nodes include multiple second-layer nodes, each of which represents a time period. Each second-layer node is connected to multiple third-layer nodes, and each third-layer node represents a light climate type. Each third-layer node is only connected to one fourth-layer node (leaf node), and each leaf node represents a light experience evaluation.

[0043] In a preferred implementation, in order to enrich the evaluation data and improve the accuracy of the scene experience evaluation in the scene experience relationship table, as shown in FIG. Figure 2 As shown, the construction process of the light scene experience relationship table also includes:

[0044] Step C1, obtaining optical data of each scenic spot under different light climate types in different time periods, where the optical data includes scenic spot illumination information and scenic spot brightness information.

[0045] In this embodiment, the scenic spot illumination information includes indoor illumination and outdoor illumination. The indoor illumination measurement principle is as follows: Figure 3 As shown, the four-corner point method or the center point method is used. During the actual measurement, the measuring points are set up starting from the indoor entrance of the scenic spot. The four-corner point method is used to measure the illumination of the entrance of the scenic spot. The illumination is measured at the four corners of the light opening, and then the average is calculated. The center point method is used to measure the illumination of the indoor floor of the scenic spot. The size of the measurement grid is generally 2m, and the grid shape is square or approximately square. Select the time period of 9:00-17:00 in a day under various light climate types for the scenic spot, and record the average illumination at the indoor entrance of the scenic spot and the illumination of multiple indoor points of the scenic spot every hour. The principle of outdoor illumination measurement is as follows: Figure 4 As shown, select an unobstructed open space around the scenic spot, and the blocking angle α formed by the light receiver and the surrounding obstructions should be less than 10° or satisfy the ratio of l to the height h of the obstruction greater than 6 times, such as Figure 4 When measuring outdoor illuminance, the photoreceiver should be placed horizontally to avoid the influence of reflected light from the ground. The outdoor ground to be measured is evenly distributed according to a 2m grid, and the illuminance of each point at each hour is measured. The average illuminance of each point per hour is calculated as the average illuminance of the area for that hour. The average of the three time periods is calculated for the whole day, that is, the outdoor average illuminance value of the node in each time period under the light climate conditions. All recorded data of indoor and outdoor illuminance measurements constitute the illuminance information of the scenic spot under different light climate types in different time periods.

[0046] Step C2, constructing a virtual model of each scenic spot, and rendering the virtual model of each scenic spot using the optical data of each scenic spot under different light climate types in different time periods to obtain a rendering model.

[0047] In this embodiment, the virtual model of each scenic spot can be constructed using the existing oblique photography method. In step C2, multiple rendering models are obtained, each rendering model is used to simulate the real scene of a certain scenic spot under a certain light climate type in a certain time period.

[0048] Step C3, obtaining scores of multiple tourists on the rendering model in multiple scene experience evaluation dimensions.

[0049] Step C4, based on the scores of multiple tourists on the rendering model in multiple light experience evaluation dimensions, obtain the virtual light experience evaluation of each scenic spot in different time periods and different light climate types. Preferably, the scores of each tourist on the multiple light experience evaluation dimensions of the rendering model are weighted and summed according to the weight of each light experience evaluation dimension to obtain the tourist score of each rendering model, and then the tourist scores of all tourists on each rendering model are averaged to obtain the virtual light experience evaluation of each rendering model, that is, the virtual light experience evaluation of each scenic spot in different time periods and different light climate types is obtained.

[0050] Step C5, combining the measured light scene experience evaluation and the virtual light scene experience evaluation of each scenic spot in different light and climate types at different time periods to construct a light scene experience relationship table. Preferably, the measured light scene experience evaluation and the virtual light scene experience evaluation of each scenic spot in different light and climate types at different time periods are weightedly summed to obtain the light scene experience evaluation of each scenic spot in different light and climate types at different time periods, and the light scene experience relationship table can be constructed according to the method in the above step B4.

[0051] In this embodiment, in order to accurately obtain the brightness information of the scenic spot, refer to Figure 5 As shown, the process of obtaining the brightness information of each scenic spot under different light climate types at different time periods includes:

[0052] Step D1, taking multiple original pictures of the scenic spot at different exposures under a preset viewing angle, wherein multiple original pictures are provided with multiple reference points at the same position. The preset viewing angle may be the main viewing angle for tourists to visit the scenic spot, such as shooting from a face-to-face viewing angle. Figure 5 In the example shown, 7 original images with different exposure intensities such as overexposure, normal exposure and underexposure are obtained under a preset viewing angle. Four reference points are set at the center of the multiple original images, and the four reference points are distributed in a matrix.

[0053] Step D2, using Photoshop software to combine multiple original images into HDR images. HDR images refer to high dynamic range images, which can provide images with a wider dynamic range and richer details than ordinary images, and can better reflect the visual effects in the real environment.

[0054] Step D3, importing the HDR image into an HDR image editing and analysis tool for Falsecolor analysis to generate a brightness false color map; the HDR image editing and analysis tool is preferably but not limited to Hdrscope. Falsecolor analysis mainly highlights the brightness information of the HDR image by artificially giving different colors to the HDR image.

[0055] Step D4, obtaining the average brightness of multiple reference points in multiple original images, recorded as the first average brightness; obtaining the average brightness of multiple reference points in the brightness pseudo-color image, recorded as the second average brightness; comparing the consistency of the second average brightness with the first average brightness to obtain a brightness comparison result;

[0056] Step D5, continuously adjust the exposure of the HDR image in Photoshop according to the brightness comparison result until the second brightness mean is consistent with the first brightness mean (consistent can be the same, or the difference between the two can be within an allowable range), and use the brightness of the adjusted HDR image and multiple reference points in the brightness pseudo-color map as the scenic spot brightness information.

[0057] In a preferred embodiment, in order to improve the accuracy of the scene experience evaluation in the modified scene experience relationship table, after the above tour recommendation method is implemented, the step of modifying the scene experience relationship table is further included:

[0058] Step E1, receiving the perception evaluations uploaded by multiple tourist terminals. Specifically, the tourist terminal may be provided with a small program or link for uploading the perception evaluations.

[0059] Step E2, obtain the average value e and standard deviation σ of multiple perceptual evaluations, and calculate the credibility Z of the perceptual evaluation X according to the following formula:

[0060]

[0061] Step E3, traverse all perceptual evaluations, if the credibility Z of the perceptual evaluation X exceeds ±3, then the perceptual evaluation X is eliminated, otherwise the perceptual evaluation X is retained; in this way, outliers can be eliminated.

[0062] Step E4, fuse all the perceptual evaluations retained after the traversal is completed with the light experience evaluations in the light experience relationship table for the corresponding time period and light climate type. The corresponding time period and light climate type can be obtained by: when tourists upload their perceptual evaluations, they will upload the transmission time synchronously, obtain the corresponding time period according to the transmission time, and obtain the corresponding light climate type from the weather forecast platform according to the transmission time. Or the acquisition method can also be: tourists synchronously upload the time period of the tour and the light climate type at that time. The specific fusion process can be: calculate the average value or weighted average value of all the perceptual evaluations retained and the light experience evaluations in the corresponding time period and light climate type in the light experience relationship table, and use the average value or weighted average value as the updated value of the light experience evaluation in the corresponding time period and light climate type in the light experience relationship table.

[0063] In this embodiment, it is further preferred that when the tourist terminal uploads the perception evaluation, the tourist terminal (such as a smart phone or a smart tablet) also uses the software Light Meter photometer to measure the brightness data of direct sunlight at the location of the smart phone in the incident light metering mode (the front camera of the smart phone is perpendicular to the incident light source), and uploads a screenshot of the brightness data display screen or the brightness data as the brightness information of the new scenic spot. After obtaining the brightness information of the new scenic spot, the corresponding leaf node is found from the light scene experience relationship table according to the current time period and current light climate type determined above, and the original scenic spot brightness information associated with the leaf node is replaced with the new scenic spot brightness information, so as to facilitate the subsequent rendering of the virtual model of the scenic spot.

[0064] This implementation method is used to correct the scenery experience evaluation in the scenery experience relationship table after the authenticity of the tourists is authenticated. Therefore, with the increase of users and the increase of operation time, the accuracy of the scenery experience relationship table is continuously improved, and the accuracy of the recommendation of the optimal tour route is continuously improved, thereby improving user satisfaction and recommendation professionalism.

[0065] In a preferred embodiment, in order to improve the efficiency of human resource scheduling in the scenic area and form daily and emergency management modes for different weather conditions in advance, the above method also includes predicting the flow of people in the scenic area according to the optimal tour route, including:

[0066] Step F1, obtaining the average value of the historical flow of people at each scenic spot in the optimal tour route under all light climate types in the predicted time period corresponding to the target tour time, and recording it as the historical flow mean.

[0067] Step F2, according to the light experience evaluation of each scenic spot in the optimal tour route at the target tour time corresponding to the predicted time period and the corresponding light climate prediction type, adjust the historical average of the flow of people at each scenic spot to obtain the predicted flow of people at each scenic spot at the target tour time. Specifically, the light experience evaluation of each scenic spot at the target tour time corresponding to the predicted time period and the corresponding light climate prediction type can be linearly mapped to a numerical interval, and the numerical interval can be [1,2]. The mapped value is used as a weight, and the historical average of the flow of people at each scenic spot is weighted by the weight to obtain the predicted flow of people at each scenic spot at the target tour time.

[0068] Step F3, outputting the predicted flow of people at each scenic spot in the optimal tour route at the target tour time.

[0069] In a preferred embodiment, in order to automatically obtain the optimal tour route with high accuracy, in step S3, determining the optimal tour route under the target tour time includes:

[0070] Step A1, initialize the pheromone concentration of the path segments between scenic spots. Take the scenic spots as nodes. The entrance and exit of the scenic spot are both special nodes, and their light experience evaluation is 0.

[0071] Step A2, performing iterative search until the number of searches reaches a preset maximum number of searches T, the t-th search process includes:

[0072] Step a: place all ants at the entrance of the scenic spot and initialize the taboo table of each ant. The taboo table is used to store the scenic spots that each ant has passed in this search.

[0073] In step b, each ant selects the next scenic spot according to the selection probability and updates the taboo table until each ant reaches the exit of the scenic spot or enters a deadlock state; the deadlock state means that the ant is trapped in a circular path finding or cannot move forward in the process of finding the optimal path.

[0074] Step c, abandon the ants that have entered the deadlock state, and update the pheromone concentration of the path segments in the path of the ants that reach the exit of the scenic area; the pheromone concentration update formula of the path segment (i, j) in a certain path S is as follows:

[0075] τ ij (t+1)←(1-ρ)·τ ij (t)+Δτ ij

[0076] ρ represents the evaporation coefficient, which is a positive number less than 1 and is generally 0.5. ij (t) represents the pheromone concentration of the path segment (i, j) in the path S in the tth search, τ ij(t + 1) represents the pheromone concentration of the path segment (i, j) in the path S at the (t + 1)-th search; Δτ ij represents the pheromone concentration increment of the path segment (i, j) in the path S:

[0077]

[0078] where Q is a constant used to adjust the scale of the pheromone increment.

[0079] Step d, if t < T, then enter the (t + 1)-th search; if t ≥ T, then enter Step A3;

[0080] Step A3, calculate the fitness of all the passing paths of the ants that reach the scenic spot exit obtained from the T searches, and select the passing path with the maximum fitness as the optimal tour route for the target tour time;

[0081] where the fitness calculation formula of the passing path S is:

[0082]

[0083] i and j respectively represent the scenic spot indexes, i ≠ j, k represents the number of scenic spots on the passing path S, E(i, t g ) represents the scenery experience evaluation of scenic spot i under the predicted time period and light climate prediction type corresponding to the target tour time t g (i, j) represents a path segment in the passing path S, D(i, j) represents the travel distance of the path segment (i, j), and T is a positive integer.

[0084] In this embodiment, further preferably, in Step A2, the selection probability calculation formula for each ant to select the next scenic spot j at scenic spot i is:

[0085]

[0086] where i, j, and m are all scenic spot indexes, i ≠ j, allowed represents the set of all scenic spots in the scenic area except the scenic spots in the taboo list of each ant, and j, m ∈ allowed; τ ij represents the pheromone concentration of the path segment from scenic spot i to scenic spot j, τ im represents the pheromone concentration of the path segment from scenic spot i to scenic spot m; δ ij represents the heuristic function for the ant to select scenic spot j as the next scenic spot at scenic spot i, δ im represents the heuristic function for the ant to select scenic spot m as the next scenic spot at scenic spot i, α represents the first adjustment parameter, generally taking values between 1 and 2; β represents the second adjustment parameter, generally taking values between 2 and 5.

[0087] This implementation is based on an intelligent bionic algorithm, combined with real-time light scene experience data and the distance between scenic spots, to generate an optimal tour route for users, thereby improving the accuracy and applicability of the optimal tour route.

[0088] In a preferred embodiment, in step S3, determining the optimal tour route under the target tour time includes:

[0089] Step G1, enumerate all candidate paths, each candidate path starts from the entrance of the scenic area, ends at the exit of the scenic area, and passes through k scenic spots without repetition, k∈[1,n], n represents the number of scenic spots in the scenic area;

[0090] Step G2, calculating the objective function value of each candidate path, and selecting the candidate path with the largest objective function value as the optimal tour route for the target tour time;

[0091] Among them, the objective function formula is:

[0092]

[0093] S' represents the candidate path, i and j represent the index of the scenic spot on the candidate path S', E(i,t g ) indicates that attraction i is within the target visiting time t g The light experience evaluation under the corresponding forecast time period and light climate forecast type; (i, j) represents a path segment in the passing path S', D(i, j) represents the travel distance of the path segment (i, j); T i represents the estimated stay time at scenic spot i (i.e., sightseeing time), and the estimated stay time is obtained by linearly mapping the average sightseeing time of scenic spot i to a dimensionless numerical interval; γ represents the first weight coefficient, and μ represents the second weight coefficient

[0094] In this embodiment, during the path generation process, the objective function is used to balance walking time, sightseeing time and light experience. This embodiment combines real-time light experience data, distance between scenic spots and sightseeing time to generate the best tour path for users. The system also selects the data uploaded by users through a credibility calculation method to ensure the accuracy and applicability of the recommended path.

[0095] The present invention also discloses a tour recommendation device based on human light scene perception, which is used to implement the above-mentioned tour recommendation method based on human light scene perception. In a preferred embodiment, the device includes:

[0096] The data acquisition module obtains the predicted time period and light climate prediction type corresponding to the target tour time, and obtains the travel distance representation between scenic spots in the scenic area;

[0097] A query module, which obtains the light experience evaluation of the scenic spot under the forecast time period and light climate forecast type from the light experience relationship table;

[0098] The module for determining the optimal tour route determines the optimal tour route under the target tour time by combining the light experience evaluation of the scenic spot and the travel distance expression between the scenic spots. The optimal tour route can balance the light experience evaluation and the travel distance expression. The optimal tour route passes through different scenic spots and starts from the entrance of the scenic spot and ends at the exit of the scenic spot.

[0099] Recommendation module recommends the best tour routes to tourists.

[0100] In this embodiment, the data acquisition module, the query module, the module for determining the optimal tour route and the recommendation module correspond one-to-one to step S1, step S2, step S3 and step S4 in the above method, and will not be described in detail here.

[0101] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", "an implementation", "a preferred implementation" or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0102] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

Claims

1. A tour recommendation method based on human scene perception, characterized in that: include: Get the predicted time period and light climate prediction type corresponding to the target tour time, and get the travel distance representation between scenic spots in the scenic area; Obtain the light experience evaluation of the scenic spot under the forecast time period and light climate forecast type from the light experience relationship table; Combine the visual experience evaluation of the scenic spot and the travel distance between the scenic spots to determine the optimal tour route under the target tour time, where the optimal tour route passes through different scenic spots and starts from the entrance of the scenic spot and ends at the exit of the scenic spot; Recommend the best tour routes to tourists.

2. A method for recommending sightseeing based on human scene perception as claimed in claim 1, characterized in that: The process of constructing the light scene experience relationship table includes: Construct multiple evaluation dimensions of light scene experience; Obtain the scores of multiple tourists on each light experience evaluation dimension of each scenic spot under different light climate types at different time periods; Based on the scores of multiple tourists on each light scene experience evaluation dimension of each scenic spot under different light climate types at different time periods, the measured light scene experience evaluation of each scenic spot under different light climate types at different time periods is obtained; The light experience relationship table is constructed using the measured light experience evaluation of each scenic spot in different time periods and different light climate types.

3. A method for recommending sightseeing based on human scene perception as claimed in claim 2, characterized in that: The process of constructing the light scene experience relationship table also includes: Acquire optical data of each scenic spot under different light climate types in different time periods, wherein the optical data includes scenic spot illumination information and scenic spot brightness information; Constructing a virtual model of each scenic spot, and rendering the virtual model of each scenic spot using optical data of each scenic spot under different light climate types at different time periods to obtain a rendering model; Obtain scores of multiple tourists on the rendering model in multiple light experience evaluation dimensions; Based on the scores of multiple tourists on the rendering model in multiple light experience evaluation dimensions, the virtual light experience evaluation of each scenic spot in different time periods and different light climate types is obtained; A light experience relationship table is constructed by combining the measured light experience evaluation and the virtual light experience evaluation of each scenic spot in different light and climate types at different time periods.

4. A method for recommending sightseeing based on human scene perception as claimed in claim 3, characterized in that: The process of obtaining the brightness information of each scenic spot under different light climate types at different time periods includes: Taking multiple original pictures of the scenic spot at different exposures at a preset viewing angle, wherein multiple reference points are set at the same position in the multiple original pictures; Use photoshop software to combine multiple original images into HDR images; Import the HDR image into the HDR image editing and analysis tool for Falsecolor analysis to generate a brightness false color map; Obtain the average brightness of multiple reference points in multiple original images, recorded as the first average brightness; obtain the average brightness of multiple reference points in the brightness pseudo-color image, recorded as the second average brightness; compare the consistency of the second average brightness with the first average brightness to obtain a brightness comparison result; According to the brightness comparison results, the exposure of the HDR image is continuously adjusted in Photoshop software until the second brightness mean is consistent with the first brightness mean, and the brightness of the adjusted HDR image and multiple reference points in the brightness pseudo-color map is used as the brightness information of the scenic spot.

5. The method for recommending sightseeing based on human scene perception as claimed in claim 3, characterized in that: It also includes the steps of modifying the light scene experience relationship table: Receive perception evaluations uploaded by multiple tourist terminals; Obtain the average value e and standard deviation σ of multiple perceptual evaluations, and calculate the credibility Z of the perceptual evaluation X according to the following formula: Traverse all perceptual evaluations. If the credibility Z of the perceptual evaluation X exceeds ±3, then eliminate the perceptual evaluation X; otherwise, retain the perceptual evaluation X. Fuse all the retained perceptual evaluations after traversal with the scene experience evaluations corresponding to the corresponding time period and light climate type in the scene experience relationship table.

6. The method for recommending sightseeing based on human scene perception as claimed in claim 1, characterized in that: It also includes predicting the scenic area passenger flow according to the optimal tour route: Obtain the average value of the historical passenger flow of each scenic spot in the optimal tour route at the predicted time period corresponding to the target tour time under all light climate types, denoted as the historical passenger flow average value. Adjust the historical passenger flow average value of each scenic spot according to the scene experience evaluation of each scenic spot in the predicted time period corresponding to the target tour time and the corresponding light climate prediction type to obtain the predicted passenger flow of each scenic spot at the target tour time. Output the predicted passenger flow of each scenic spot in the optimal tour route at the target tour time.

7. A method for recommending sightseeing based on human scene perception as claimed in any one of claims 1 to 6, characterized in that: The determination of the optimal tour route at the target tour time includes: Step A1, initialize the pheromone concentration of the path segments between scenic spots. Step A2, perform iterative search until the search times reach the preset maximum search times T. The t-th search process includes: Release all ants at the scenic area entrance and initialize the taboo list of each ant. Each ant selects the next scenic spot according to the selection probability and updates the taboo list until each ant reaches the scenic area exit or enters a deadlock state. Abandon the ants that enter the deadlock state and update the pheromone concentration of the path segments in the passing paths of the ants that reach the scenic area exit. If t < T, then enter the (t + 1)-th search. If t ≥ T, then enter Step A3. Step A3, calculate the fitness of the passing paths of all ants that reach the scenic area exit obtained from T searches, and select the passing path with the maximum fitness as the optimal tour route at the target tour time. Among them, the fitness calculation formula of the passing path S is: i, j represent the index of scenic spots respectively, i≠j, k represents the number of scenic spots on the path S, E(i,t g ) indicates that attraction i is within the target visiting time t g The light experience evaluation under the corresponding forecast time period and light climate forecast type; (i, j) represents a path segment in the passing path S, D(i, j) represents the travel distance of the path segment (i, j), and T is a positive integer.

8. The method for recommending sightseeing based on human scene perception as claimed in claim 7, characterized in that: In Step A2, the selection probability calculation formula for each ant to select the next scenic spot j at scenic spot i is: Among them, i, j, m are all scenic spot indexes, i≠j, allowed represents the set of all scenic spots in the scenic area except the scenic spots in the taboo table of each ant, j, m∈allowed; τ ij represents the pheromone concentration of the path segment from attraction i to attraction j, τ im represents the pheromone concentration of the path segment from scenic spot i to scenic spot m; δ ij represents the heuristic function for ants to choose attraction j as the next attraction at attraction i, δ im represents the heuristic function for ants to select attraction m as the next attraction at attraction i; α represents the first adjustment parameter, and β represents the second adjustment parameter.

9. A method for recommending sightseeing based on human scene perception as claimed in any one of claims 1 to 6, characterized in that: The determination of the optimal tour route at the target tour time includes: Enumerate all candidate paths. Each candidate path starts from the scenic area entrance, ends at the scenic area exit, and passes through k scenic spots without repetition, where k ∈ [1, n], and n represents the number of scenic spots in the scenic area. Calculate the objective function value of each candidate path, and select the candidate path with the maximum objective function value as the optimal tour route at the target tour time. Among them, the objective function formula is: S' represents the candidate path, i and j represent the index of the scenic spot on the candidate path S', E(i,t g ) indicates that attraction i is within the target visiting time t g The light experience evaluation under the corresponding forecast time period and light climate forecast type; (i, j) represents a path segment in the passing path S', D(i, j) represents the travel distance of the path segment (i, j); T i represents the estimated stay time at scenic spot i; γ represents the first weight coefficient, and μ represents the second weight coefficient.

10. A tour recommendation device based on human scene perception, characterized in that: A method for tour recommendation based on human perception of light and scene, which is used to implement one of claims 1-9, includes: A data acquisition module, which acquires the predicted time period corresponding to the target tour time and the light climate prediction type, and acquires the representation of the travel distance between scenic spots in the scenic area. A query module, which acquires the scene experience evaluations of scenic spots under the predicted time period and the light climate prediction type from the scene experience relationship table. An optimal tour route determination module, which combines the scene experience evaluations of scenic spots and the representation of the travel distance between scenic spots to determine the optimal tour route at the target tour time. The optimal tour route passes through different scenic spots, starts from the scenic area entrance, and ends at the scenic area exit. A recommendation module, which recommends the optimal tour route to tourists.

Images