A digital human modeling method and system for cultural tourism

Abstract

The application relates to the technical field of digital human generation, and particularly discloses a cultural and travel digital human modeling method and system, which comprises the following steps: acquiring scene information of a cultural and travel scene, and creating a digital twin model; acquiring a video of a public area, performing feature recognition on the video, constructing a virtual digital human, and inserting the virtual digital human into the digital twin model; receiving an access request with an access parameter sent by a remote user, intercepting the digital twin model with the virtual digital human according to the access parameter, and sending the digital twin model to the remote user; acquiring feedback information of the remote user in real time, and adjusting the intercepted digital twin model with the virtual digital human according to the feedback information. The remote tourism service provided by the application is equivalent to converting a single-player game into a network game, greatly improves the authenticity, helps to improve the utilization rate, and provides good cultural and travel services for people who are inconvenient to travel.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of digital human generation, and particularly relates to a cultural and tourism digital human modeling method and system. BACKGROUND

[0002] The cultural and tourism digital human is a short form of cultural and tourism digital human, and is an innovative application of the cultural and tourism industry in combination with digital technology, artificial intelligence and virtual reality and other frontier technologies in recent years. It mainly refers to a virtual character created by digital modeling, AI driving, speech synthesis, motion capture and other technologies, and is used to improve the interactivity, immersion and dissemination effect of cultural and tourism scenes.

[0003] The existing cultural and tourism digital human is only used to provide explanation services to assist guide personnel in work and travel, and the interaction position of most cultural and tourism digital humans is fixed, which causes the problems of insufficient interactivity and low utilization rate. Therefore, how to improve the interactivity and utilization rate of the cultural and tourism digital human is a technical problem to be solved by the technical scheme of the present application. SUMMARY

[0004] The present application aims to provide a cultural and tourism digital human modeling method and system to solve the problems in the background art.

[0005] To achieve the above-mentioned purpose, the present application provides the following technical scheme:

[0006] A cultural and tourism digital human modeling method and system, the method comprising:

[0007] Obtaining scene information of a cultural and tourism scene, and creating a digital twin model;

[0008] Obtaining a video of a public area, performing feature recognition on the video, constructing a virtual digital human, and inserting the virtual digital human into the digital twin model;

[0009] Receiving an access request containing an access parameter sent by a remote user, intercepting a digital twin model containing a virtual digital human according to the access parameter, and sending the digital twin model to the remote user;

[0010] During the process of sending the intercepted digital twin model containing the virtual digital human to the remote user, real-time feedback information of the remote user is obtained, and the intercepted digital twin model containing the virtual digital human is adjusted according to the feedback information.

[0011] As a further scheme of the present application, the step of obtaining scene information of a cultural and tourism scene and creating a digital twin model comprises:

[0012] Obtaining a map of the cultural and tourism scene, and positioning a passable area according to the block features of the map;

[0013] Obtaining a center line of the passing area, selecting a middle point on the center line based on a preset step length;

[0014] Obtaining a main body level of each middle point in a direction perpendicular to the center line, and determining an offset of the middle point according to the main body level;

[0015] Determining an offset point of each middle point based on the offset;

[0016] Generating a motion path based on the offset point, sending the motion path to a vision acquisition robot, and obtaining vision information; wherein the process of obtaining the vision information is to rotate to acquire images at each offset point, and the acquisition direction rotates 360 degrees in the horizontal plane;

[0017] Creating a digital twin model based on the vision information and a map of the travel scene;

[0018] The calculation process of the offset is as follows:

[0019] In the formula, P represents the offset, a is a preset correction coefficient, N is the number of all main bodies in the direction perpendicular to the center line, L i represents the main body level of the i-th main body, d i represents the distance between the i-th main body and the middle point, the distance has a sign, and the sign is determined by the directional relationship between the main body and the middle point. The positive sign corresponds to the preset value.

[0020] As a further scheme of the present application: the step of obtaining a video of the public area, performing feature recognition on the video, constructing a virtual digital person, and inserting the digital twin model includes:

[0021] Obtaining a video of the public area;

[0022] Tracking the personnel in the video, determining the personnel frame at each time, and synchronously determining the personnel position;

[0023] When detecting the personnel frame, performing contour recognition on the area within the personnel frame, and dividing the personnel frame into area blocks; wherein the variance of pixel values in each area block is less than a preset variance threshold;

[0024] Constructing a virtual digital person based on the area block;

[0025] Inserting the virtual digital person into the digital twin model according to the personnel position to obtain a digital twin model containing the virtual digital person at each time.

[0026] As a further scheme of the present application: the step of constructing a virtual digital person based on the area block includes:

[0027] The statistical segmentation result and the corresponding virtual digital person are used to construct a mapping table; the mapping table includes a segmentation result item and a virtual digital person item, the segmentation result is a matrix generated by the region block, when the personnel position corresponding to the region block does not belong to the travel scene, the data item corresponding to the region block is deleted in the mapping table;

[0028] When a new segmentation result is obtained, the virtual digital person is matched in the mapping table;

[0029] When the matching fails, the region block obtained by segmentation is read in sequence, the position positioning model library in the personnel frame is read, and the model is queried in the model library based on the region block;

[0030] The queried model is counted according to the position sequence of the region block in the personnel frame, the virtual digital person is constructed, and the mapping table is updated synchronously;

[0031] The generation process of the segmentation result is as follows:

[0032] The average color value of each region block is obtained, and the average color value is counted according to the position correlation of the region block to obtain a matrix as the segmentation result.

[0033] As a further scheme of the present application: the step of receiving the access request containing the access parameter sent by the remote user, and intercepting the digital twin model containing the virtual digital person according to the access parameter, and sending to the remote user includes:

[0034] Receiving the access request sent by the remote user, and opening the parameter acquisition port to the user;

[0035] Based on the parameter acquisition port, the access time, access coordinates and access direction input by the remote user are received;

[0036] According to the access time, the digital twin model is positioned, and the intercepting map is obtained according to the access coordinates and the access direction;

[0037] Based on the intercepting map, the digital twin model containing the virtual digital person is intercepted and sent to the remote user.

[0038] As a further scheme of the present application: the step of adjusting the intercepted digital twin model containing the virtual digital person according to the feedback information of the remote user in the process of sending the intercepted digital twin model containing the virtual digital person to the remote user includes:

[0039] In the process of sending the intercepted digital twin model containing the virtual digital person to the remote user, the face data of the remote user is obtained in real time based on the pre-granted interaction permission;

[0040] The face data is input into the trained emotion evaluation model, and the degree of favor is output;

[0041] query the virtual digital human in the intercepted digital twin model, as the remote user's favorability to the virtual digital human;

[0042] adjust all virtual digital humans in the favorability digital twin model.

[0043] The technical scheme of the present application also provides a cultural and travel digital human modeling system, the system comprises:

[0044] a twin model creation module for obtaining scene information of a cultural and travel scene and creating a digital twin model;

[0045] a digital human construction module for obtaining a video of a public area, performing feature recognition on the video, constructing a virtual digital human, and inserting the virtual digital human into the digital twin model;

[0046] a model interception and display module for receiving an access request with an access parameter sent by a remote user, intercepting a digital twin model with a virtual digital human according to the access parameter, and sending the digital twin model to the remote user;

[0047] a display content adjustment module for obtaining feedback information of the remote user in real time during the process of sending the intercepted digital twin model with the virtual digital human to the remote user, and adjusting the intercepted digital twin model with the virtual digital human according to the feedback information.

[0048] As a further scheme of the present application, the twin model creation module comprises:

[0049] a passable area positioning unit for obtaining a map of the cultural and travel scene and positioning a passable area according to block features of the map;

[0050] an intermediate point selection unit for obtaining a center line of the passable area and selecting an intermediate point based on a preset step length on the center line;

[0051] an offset amount determination unit for obtaining a main body level of each intermediate point in a direction perpendicular to the center line and determining an offset amount of the intermediate point according to the main body level;

[0052] an offset point determination unit for determining an offset point of each intermediate point based on the offset amount;

[0053] a visual information acquisition unit for generating a motion path based on the offset point and sending the motion path to a visual acquisition robot to obtain visual information; wherein the process of obtaining the visual information is to rotate to collect images at each offset point, and the collection direction rotates 360 degrees in the horizontal plane;

[0054] a creation execution unit for creating a digital twin model based on the visual information and the map of the cultural and travel scene;

[0055] The calculation process of the offset is as follows:

[0056] In the formula, P represents the offset, alpha is a preset correction coefficient, N is the number of all subjects in the vertical median direction, L i represents the subject level of the i-th subject, d i represents the distance between the i-th subject and the intermediate point, the distance has a sign, and the sign is determined by the directional relationship between the subject and the intermediate point. The positive sign corresponds to a preset direction.

[0057] As a further scheme of the present application, the digital human construction module comprises:

[0058] A video acquisition unit is configured to acquire a video of a public area.

[0059] A personnel positioning unit is configured to perform personnel tracking on the video to determine personnel frames at each time and synchronously determine personnel positions.

[0060] A frame cutting unit is configured to, when a personnel frame is detected, perform contour recognition on a region within the personnel frame and cut the personnel frame into region blocks, wherein a variance of pixel values in each region block is less than a preset variance threshold.

[0061] A region block application unit is configured to construct a virtual digital human based on the region blocks.

[0062] An insertion execution unit is configured to insert the virtual digital human into a digital twin model according to the personnel positions to obtain a digital twin model containing the virtual digital human at each time.

[0063] As a further scheme of the present application, the model interception display module comprises:

[0064] A port opening unit is configured to receive an access request sent by a remote user and open a parameter acquisition port to the user.

[0065] A parameter receiving unit is configured to receive an access time, an access coordinate and an access direction input by the remote user based on the parameter acquisition port.

[0066] A map acquisition unit is configured to locate the digital twin model according to the access time and acquire an intercepted map according to the access coordinate and the access direction.

[0067] An interception execution unit is configured to intercept the digital twin model containing the virtual digital human based on the intercepted map and send the digital twin model to the remote user.

[0068] Compared with the prior art, the present application has the following beneficial effects:

[0069] The application constructs a digital twin model according to a travel scene, acquires tourists according to public videos, constructs a virtual digital person corresponding to the tourists in real time, inserts the virtual digital person into the digital twin model, and provides remote travel services based on the digital twin model containing the virtual digital person with real-time changes, so that the virtual digital person has high interactivity and high restoration, and the single-player game is converted into a network game, the authenticity is greatly improved, the interactivity and the use rate of the existing travel digital person are improved, and good travel services are provided for people who are inconvenient to travel. BRIEF DESCRIPTION OF DRAWINGS

[0070] In order to more clearly illustrate the technical solutions in the embodiments of the application, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the application.

[0071] Figure 1 A total flow block diagram of a travel digital person modeling method is shown.

[0072] Figure 2 A structure diagram of a travel digital person modeling system is shown. DETAILED DESCRIPTION

[0073] In order to make the technical problems to be solved by the application, the technical solutions and the beneficial effects more clearly, the following will further describe the application in combination with the drawings and the embodiments. It should be understood that the specific embodiments described herein are only used to explain the application, and are not used to limit the application.

[0074] Figure 1 A total flow block diagram of a travel digital person modeling method and system, in the embodiments of the application, a travel digital person modeling method, the method comprises:

[0075] Step S100: acquiring scene information of a travel scene, and creating a digital twin model;

[0076] The technical scheme of the application is applied to an area available for touring, such as a scenic spot, which is referred to as a travel scene, a map and visual information of the travel scene are acquired, a three-dimensional model is constructed based on a preset scale, and the three-dimensional model is referred to as a digital twin model because the construction process needs to update data in a timely manner.

[0077] Step S200: acquiring a video of a public area, performing feature recognition on the video, constructing a virtual digital person, and inserting the virtual digital person into the digital twin model;

[0078] The travel scene generally belongs to a public area, a camera is installed in the public area, a video of the public area is acquired in real time by the camera, feature recognition is performed on the video, a virtual digital person corresponding to the tourist can be obtained, since the positions of the tourists at different moments are different, the virtual digital persons at different moments are also different, and after the digital twin model is inserted, the digital twin model at different moments is obtained.

[0079] It should be noted that although the video in the public area is acquired by default with the permission, the process of processing the video to generate the virtual digital person needs to be informed to the tourist, and after obtaining the explicit authorization of the tourist, the video can be acquired and the video can be "secondarily created".

[0080] Step S300: receiving an access request containing an access parameter sent by a remote user, intercepting a digital twin model containing a virtual digital person according to the access parameter, and sending the digital twin model to the remote user;

[0081] The technical scheme of the present application constructs a digital twin model containing a virtual digital person, provides remote tour service, receives an access request containing an access parameter sent by a remote user, the access parameter indicates which content is accessed by the remote user at which time, intercepts a digital twin model containing a virtual digital person according to the access parameter, and sends the digital twin model to the remote user.

[0082] Specifically, as for the remote access process, the moment when the remote user sends the access request is taken as the access moment, after sending the access request, the execution subject of the technical scheme of the present application establishes a connection channel with the remote user, receives the perspective (one kind of access parameter) selected by the remote user on the personal device, intercepts the digital twin model containing the virtual digital person in the perspective, and sends the digital twin model to the remote user.

[0083] Step S400: in the process of sending the intercepted digital twin model containing the virtual digital person to the remote user, real-time feedback information of the remote user is acquired, and the intercepted digital twin model containing the virtual digital person is adjusted according to the feedback information;

[0084] After sending the intercepted digital twin model containing the virtual digital person to the remote user, the feedback information sent by the remote user is received in real time, the simplest feedback information is the preference degree of each virtual digital person input by the remote user, the virtual digital person with higher preference degree is replaced by the virtual digital person with lower preference degree, and the digital twin model more liked by the remote user can be obtained, and the remote tour experience of the remote user is optimized.

[0085] As for step S100, the step of acquiring the scene information of the travel scene and creating the digital twin model comprises:

[0086] Obtain a map of the travel scene, locate a passable area according to a plot feature of the map;

[0087] Obtain a center line of the passable area, and select a middle point on the center line based on a preset step length;

[0088] Obtain a main body level of each middle point in a direction perpendicular to the center line, and determine an offset of the middle point according to the main body level;

[0089] Determine an offset point of each middle point based on the offset;

[0090] Generate a motion path based on the offset point, send the motion path to a visual collection robot, and obtain visual information; wherein, the process of obtaining the visual information is to rotate and collect images at each offset point, and the collection direction rotates 360 degrees in a horizontal plane;

[0091] Create a digital twin model based on the visual information and the map of the travel scene.

[0092] Obtain a map of the travel scene, locate a passable area according to a plot feature of the map, the existing map service is very developed, the positioning process of the passable area is very easy, and the passable area in the map provided by the management party of the travel scene is very obvious and accurate; obtain a center line of the passable area, and select a middle point on the center line based on a preset step length, the step length refers to selecting a point on the center line every certain distance, called a middle point, obtain a main body level of each middle point in a direction perpendicular to the center line, and determine an offset of the middle point according to the main body level, determine an offset point of each middle point based on the offset, take all offset points corresponding to the middle points as passing points, generate a motion path, send the motion path to a visual collection robot, and obtain visual information; wherein, the visual collection robot continuously collects images by rotating 360 degrees every time it reaches an offset point, and obtains visual information, the collection direction during the rotation process is in the same horizontal plane (or in a horizontal plane interlayer with a very low height, with a vertical deviation less than a preset threshold); create a digital twin model based on the visual information and the map of the travel scene.

[0093] Among them, for the determination process of the offset, there are actually two directions of each middle point perpendicular to the center line, located on both sides of the center line, obtain the main bodies on both sides of the center line in the travel scene, query the main body level of the main bodies, the main body level is existing data, using the rating of the scenic spot, the offset can be determined according to the main body level.

[0094] The calculation process of the offset includes:

[0095] In the formula, P represents the offset, a is a preset correction coefficient, N is the number of all main bodies in the direction perpendicular to the center line, L ia subject level representing an i-th subject, d i a distance between the i-th subject and the intermediate point, the distance having a sign, the sign being determined by a directional relationship between the subject and the intermediate point, a positive sign corresponding to a preset direction.

[0096] As to the direction corresponding to the positive sign, either direction on the two sides can be set as positive, and the calculated value is positive when the influence of all subjects considered in the above calculation process is positive, and the calculated value is negative when the influence of all subjects considered in the above calculation process is negative.

[0097] As to step S200, the steps of obtaining a video of a public area, performing feature recognition on the video, constructing a virtual digital person, and inserting the digital twin model include:

[0098] obtaining a video of a public area;

[0099] performing personnel tracking on the video to determine personnel frames at each time and synchronously determine personnel positions;

[0100] when a personnel frame is detected, performing contour recognition on a region in the personnel frame to divide the personnel frame into region blocks; wherein a variance of pixel values in each region block is less than a preset variance threshold;

[0101] constructing a virtual digital person based on the region blocks;

[0102] inserting the virtual digital person into the digital twin model according to the personnel positions to obtain digital twin models containing the virtual digital person at each time.

[0103] The video of the public area is obtained, and the personnel frames at each time can be determined by means of an existing personnel tracking algorithm. After positioning the personnel frame, the personnel position is also determined. The personnel position generally adopts the center position of the bottom edge of the personnel frame and is a coordinate. When the personnel frame is detected, the region in the personnel frame is subjected to contour recognition to divide the personnel frame into region blocks. The pixel values of each pixel point in the same region block are almost the same. After the region blocks are determined, a virtual digital person is constructed based on the region blocks. The virtual digital person is inserted into the digital twin model according to the personnel positions to obtain digital twin models containing the virtual digital person at each time.

[0104] Specifically, the step of constructing a virtual digital person based on the region blocks includes:

[0105] statistically analyzing the division results and the corresponding virtual digital persons to construct a mapping table; the mapping table includes a division result item and a virtual digital person item, and the division result is a matrix generated by the region blocks. When the personnel position corresponding to the region blocks does not belong to a travel and tourism scene, the data item corresponding to the region blocks is deleted in the mapping table.

[0106] When a new segmentation result is obtained, match the virtual digital person in the mapping table;

[0107] When the matching fails, read the region block obtained by segmentation in sequence, and locate the model library in the position of the person frame, query the model in the model library based on the region block;

[0108] According to the position sequence of the region block in the person frame, the queried model is counted to construct a virtual digital person, and the mapping table is updated synchronously.

[0109] In an example of the technical scheme of the present application, the conversion relationship between the region block and the virtual digital person is described. For a person frame, after contour recognition, a plurality of region blocks can be obtained, the average color value of each region block is obtained, and then the average color value is counted according to the position correlation of the region block to obtain a matrix (the matrix can have missing data). At this time, the matrix is taken as the segmentation result, representing the characteristics of a tourist. After each tourist constructs a virtual digital person, it is stored in the mapping table. The mapping table includes a segmentation result item and a virtual digital person item, and the segmentation result is the above-mentioned matrix. For a new tourist, after obtaining the region block, a matrix is constructed based on the region block, and the matrix is traversed in the mapping table to query whether there is a constructed virtual digital person. If there is, it can be directly read. If there is not, a new virtual digital person needs to be constructed.

[0110] Specifically, the way to construct a new virtual digital person includes:

[0111] Read the region block obtained by segmentation in sequence, and locate the model library in the position of the person frame, query the model in the model library based on the region block; count the queried model according to the position sequence of the region block in the person frame to obtain a virtual digital person. At the same time, the segmentation result (matrix) of the region block is obtained, the obtained virtual digital person is created, and a new data item is inserted into the mapping table.

[0112] In addition, the data storage time length in the mapping table can be very long, such as one month, which means that the tourists visiting the travel scene within one month will be recorded. However, the mapping table will be very large in this way, therefore, the technical scheme of the present application further provides a self-deletion function: when the position of the region block corresponding to the personnel does not belong to the travel scene, the data item corresponding to the region block is deleted in the mapping table.

[0113] Regarding step S300, the step of receiving the access request containing the access parameter sent by the remote user, and intercepting the digital twin model containing the virtual digital person according to the access parameter, and sending to the remote user includes:

[0114] Receive the access request sent by the remote user, and open the parameter acquisition port to the user;

[0115] acquire access time, access coordinates and access direction input by the remote user based on the parameter acquisition port;

[0116] position the digital twin model according to the access time, acquire the intercepted map according to the access coordinates and the access direction, and intercept the digital twin model containing the virtual digital person based on the intercepted map, and send to the remote user.

[0117] acquire access time, access coordinates and access direction input by the remote user based on the parameter acquisition port;

[0118] acquire access time, access coordinates and access direction input by the remote user based on the parameter acquisition port, position the digital twin model according to the access time, acquire the intercepted map according to the access coordinates and the access direction, intercept the digital twin model containing the virtual digital person based on the intercepted map, and send to the remote user. This process is very simple, which is equivalent to constructing a virtual three-dimensional scene, and the remote user can watch the model from different angles in the virtual three-dimensional scene. Many existing applications based on the first-person perspective have this function.

[0119] Regarding step S400, the step of acquiring feedback information of the remote user in real time during the process of sending the intercepted digital twin model containing the virtual digital person to the remote user, and adjusting the intercepted digital twin model containing the virtual digital person according to the feedback information includes:

[0120] acquiring facial data of the remote user in real time based on the pre-granted interaction permission during the process of sending the intercepted digital twin model containing the virtual digital person to the remote user;

[0121] inputting the facial data into the trained emotion evaluation model to output the degree of favor;

[0122] querying the virtual digital person in the intercepted digital twin model as the degree of favor of the remote user to the virtual digital person;

[0123] adjusting all virtual digital persons in the digital twin model based on the degree of favor.

[0124] In one example of the technical scheme of the present application, the data structure of the feedback information is specifically limited, and during the process of sending the intercepted digital twin model containing the virtual digital person to the remote user, the facial data of the remote user is acquired in real time based on the pre-granted interaction permission, the facial data is input into the trained emotion evaluation model to output the degree of favor, at this time, the obtained degree of favor is considered as the current observed degree of favor to the virtual digital person, therefore, the virtual digital person in the intercepted digital twin model is queried as the degree of favor of the remote user to the virtual digital person.

[0125] Adjustment is made based on all virtual digital humans in the likeability digital twin model, and the specific operation process includes:

[0126] The virtual digital human with the likeability greater than the preset first likeability threshold is read, the average color value of each area block corresponding to the virtual digital human is read, the average color value is counted based on the position of the area block, the adjustment image filter is constructed based on the preset transparency, and the image filter is superimposed on the virtual digital human with the likeability less than the preset second likeability threshold.

[0127] The principle of this process is that the focus of the virtual digital human is on the color, and the color value feature of the virtual digital human with high likeability is superimposed on the virtual digital human with low likeability, so as to optimize the color.

[0128] Figure 2 The structural diagram of the digital human modeling system is shown, and in an preferred embodiment of the technical scheme of the present application, a digital human modeling system for cultural tourism is also provided, the system 10 includes:

[0129] The twin model creation module 11 is used to obtain scene information of a cultural tourism scene, and create a digital twin model;

[0130] The digital human construction module 12 is used to obtain a video of a public area, perform feature recognition on the video, construct a virtual digital human, and insert the virtual digital human into the digital twin model;

[0131] The model intercepting and displaying module 13 is used to receive an access request containing an access parameter sent by a remote user, intercept the digital twin model containing the virtual digital human according to the access parameter, and send the digital twin model to the remote user;

[0132] The display content adjustment module 14 is used to obtain feedback information of the remote user in real time during the process of sending the intercepted digital twin model containing the virtual digital human to the remote user, and adjust the intercepted digital twin model containing the virtual digital human according to the feedback information.

[0133] Further, the twin model creation module 11 includes:

[0134] The passable area positioning unit is used to obtain a map of the cultural tourism scene, and position the passable area according to the block features of the map;

[0135] The intermediate point selection unit is used to obtain a center line of the passable area, and select an intermediate point on the center line based on a preset step length;

[0136] The offset determination unit is used to obtain a body level of each intermediate point in the direction perpendicular to the center line, and determine the offset of the intermediate point according to the body level;

[0137] The offset point position determination unit is configured to determine offset point positions of the intermediate point positions based on the offset amount;

[0138] The visual information acquisition unit is configured to generate a motion path based on the offset point positions, send the motion path to a visual acquisition robot, and acquire visual information; wherein the process of acquiring the visual information comprises: rotatingly acquiring images at each offset point position, and rotating the acquisition direction by 360 degrees in the horizontal plane;

[0139] The digital twin creation execution unit is configured to create a digital twin model based on the visual information and a map of the cultural and tourism scene.

[0140] Specifically, the digital human construction module 12 comprises:

[0141] The video acquisition unit is configured to acquire a video of the public area;

[0142] The personnel positioning unit is configured to perform personnel tracking on the video, determine a personnel frame at each time, and synchronously determine a personnel position;

[0143] The frame segmentation unit is configured to, when detecting the personnel frame, perform contour recognition on a region in the personnel frame, and segment the personnel frame into region blocks; wherein a variance of pixel values in each region block is less than a preset variance threshold;

[0144] The region block application unit is configured to construct a virtual digital human based on the region blocks;

[0145] The insertion execution unit is configured to insert the virtual digital human into the digital twin model according to the personnel position, to obtain a digital twin model containing the virtual digital human at each time.

[0146] Further, the model interception display module 13 comprises:

[0147] The port opening unit is configured to receive an access request sent by a remote user, and open a parameter acquisition port to the user;

[0148] The parameter receiving unit is configured to receive an access time, an access coordinate, and an access direction input by the remote user based on the parameter acquisition port;

[0149] The map acquisition unit is configured to position the digital twin model according to the access time, and acquire an intercepted map according to the access coordinate and the access direction;

[0150] The interception execution unit is configured to intercept the digital twin model containing the virtual digital human based on the intercepted map, and send the digital twin model to the remote user.

[0151] The above merely describes the preferred embodiments of the present application, and is not intended to limit the patent scope of the present application, and any equivalent structure or equivalent process conversion, or direct or indirect application in other related technical fields, which are made by using the content of the present application specification and drawings, are also included in the patent protection scope of the present application.

Claims

1. A modeling method of a digital human in a tourism and travel field, characterized in that, The method comprises: Obtaining scene information of a travel scene, and creating a digital twin model; Obtaining a video of a public area, performing feature recognition on the video, constructing a virtual digital person, and inserting the virtual digital person into the digital twin model; Receiving an access request with access parameters sent by a remote user, intercepting a digital twin model with a virtual digital person according to the access parameters, and sending the digital twin model to the remote user; During the process of sending the intercepted digital twin model with the virtual digital person to the remote user, real-time feedback information of the remote user is obtained, and the intercepted digital twin model with the virtual digital person is adjusted according to the feedback information; The step of obtaining a video of a public area, performing feature recognition on the video, constructing a virtual digital person, and inserting the virtual digital person into the digital twin model comprises: Obtaining a video of a public area; Tracking personnel in the video, determining personnel frames at each time, and synchronously determining personnel positions; When detecting a personnel frame, performing contour recognition on the area in the personnel frame, and cutting the personnel frame into area blocks; wherein the variance of pixel values in each area block is less than a preset variance threshold; Constructing a virtual digital person based on the area blocks; Inserting the virtual digital person into the digital twin model according to the personnel positions to obtain a digital twin model with a virtual digital person at each time; The step of receiving an access request with access parameters sent by a remote user, intercepting a digital twin model with a virtual digital person according to the access parameters, and sending the digital twin model to the remote user comprises: Receiving an access request sent by a remote user, and opening a parameter acquisition port to the user; Receiving access time, access coordinates, and access direction input by the remote user based on the parameter acquisition port; Positioning the digital twin model according to the access time, and obtaining an intercepted map according to the access coordinates and the access direction; Intercepting a digital twin model with a virtual digital person based on the intercepted map, and sending the digital twin model to the remote user; The step of obtaining a video of a public area, performing feature recognition on the video, constructing a virtual digital person, and inserting the virtual digital person into the digital twin model comprises: During the process of sending the intercepted digital twin model with the virtual digital person to the remote user, real-time feedback information of the remote user is obtained, and the intercepted digital twin model with the virtual digital person is adjusted according to the feedback information; During the process of sending the intercepted digital twin model with the virtual digital person to the remote user, real-time feedback information of the remote user is obtained, and the intercepted digital twin model with the virtual digital person is adjusted according to the feedback information; The step of obtaining scene information of a travel scene, and creating a digital twin model comprises: Obtaining a map of a travel scene, and positioning a passable area according to the block features of the map; During the process of sending the intercepted digital twin model with the virtual digital person to the remote user, real-time feedback information of the remote user is obtained, and the intercepted digital twin model with the virtual digital person is adjusted according to the feedback information; 2. The digital human modeling method for tourism and travel according to claim 1, wherein, ​ ​ Obtaining a center line of the passing area, selecting a middle point on the center line based on a preset step length; Obtaining a main body level of each middle point in a direction perpendicular to the center line, and determining an offset of the middle point according to the main body level; Determining an offset point of each middle point based on the offset; Generating a motion path based on the offset point, and sending the motion path to a visual acquisition robot to obtain visual information; wherein the process of obtaining the visual information is to rotate to collect images at each offset point, and the collection direction rotates 360 degrees in the horizontal plane; Creating a digital twin model based on the visual information and a map of the cultural and tourism scene; The calculation process of the offset is as follows: In the formula, Indicates the offset. The preset correction factor. This represents the number of all entities along the vertical centerline. Indicates the first The subject level of each subject, Indicates the first The distance between a subject and a midpoint, wherein the distance has a sign, the sign being determined by the directional relationship between the subject and the midpoint, and the direction corresponding to the positive sign is a preset value.

3. The digital human modeling method for tourism and travel according to claim 1, wherein, The step of constructing a virtual digital person based on the area block includes: Statistically analyzing the segmentation result and the corresponding virtual digital person to construct a mapping table; the mapping table includes a segmentation result item and a virtual digital person item, and the segmentation result is a matrix generated by the area block; when the position of the personnel corresponding to the area block does not belong to the cultural and tourism scene, the data item corresponding to the area block is deleted in the mapping table; When a new segmentation result is obtained, the virtual digital person is matched in the mapping table; When the matching fails, the area block obtained by segmentation is read in sequence, the position positioning model library in the personnel frame is read, and the model is queried in the model library based on the area block; According to the position sequence of the area block in the personnel frame, the queried model is statistically analyzed, the virtual digital person is constructed, and the mapping table is updated synchronously. The generation process of the segmentation result is as follows: Obtain the average color value of each area block, and then statistically analyze the average color value according to the position correlation of the area block to obtain a matrix as the segmentation result.

4. A digital human modeling system for cultural tourism, characterized in that, The system comprises: A twin model creation module for obtaining scene information of a cultural and tourism scene and creating a digital twin model; A digital person construction module for obtaining a video of a public area, performing feature recognition on the video, constructing a virtual digital person, and inserting the virtual digital person into the digital twin model; A model interception display module for receiving an access request containing an access parameter sent by a remote user, intercepting a digital twin model containing a virtual digital person according to the access parameter, and sending the digital twin model to the remote user; A display content adjustment module for obtaining feedback information of the remote user in real time during the process of sending the intercepted digital twin model containing the virtual digital person to the remote user, and adjusting the intercepted digital twin model containing the virtual digital person according to the feedback information; The digital person construction module comprises: A video acquisition unit for obtaining a video of a public area; A personnel positioning unit for tracking personnel in the video and determining personnel frames at each time point and personnel positions synchronously; A frame segmentation unit for performing contour recognition on an area in a personnel frame when the personnel frame is detected, and segmenting the personnel frame into area blocks; wherein the variance of pixel values in each area block is less than a preset variance threshold; An area block application unit for constructing a virtual digital person based on the area block; An insertion execution unit for inserting the virtual digital person into the digital twin model according to the personnel position to obtain a digital twin model containing the virtual digital person at each time point; The model interception display module comprises: A port opening unit is configured to receive an access request sent by a remote user and open a parameter acquisition port for the user. A parameter receiving unit is configured to receive an access time, an access coordinate and an access direction input by the remote user based on the parameter acquisition port. A map acquisition unit is configured to locate a digital twin model according to the access time and acquire a cut map according to the access coordinate and the access direction. A cutting execution unit is configured to cut the digital twin model containing a virtual digital person based on the cut map and send the digital twin model to the remote user. The working content of the display content adjustment module includes: In the process of sending the cut digital twin model containing the virtual digital person to the remote user, the facial data of the remote user is acquired in real time based on the pre-granted interaction permission; The facial data is input into the trained emotion evaluation model, and the degree of favor is output; The virtual digital person in the cut digital twin model is queried as the degree of favor of the remote user to the virtual digital person; Based on the degree of favor, all virtual digital persons in the digital twin model are adjusted; the specific operation process includes: Read the virtual digital person whose degree of favor is greater than a preset first degree of favor threshold, read the average color value of each region block corresponding to the virtual digital person, count the average color value based on the position of the region block, construct an adjustment image filter based on a preset transparency, and superimpose the image filter on the virtual digital person whose degree of favor is less than a preset second degree of favor threshold.

5. The digital human modeling system of claim 4, wherein, The twin model creation module includes: A pass-through area positioning unit is configured to acquire a map of a travel and tourism scene and locate a pass-through area according to a block feature of the map. An intermediate point selection unit is configured to acquire a center line of the pass-through area and select an intermediate point based on a preset step length on the center line. An offset amount determination unit is configured to acquire a main body level of each intermediate point in a direction perpendicular to the center line and determine an offset amount of the intermediate point according to the main body level. An offset point determination unit is configured to determine an offset point of each intermediate point based on the offset amount. A visual information acquisition unit is configured to generate a motion path based on the offset point, send the motion path to a visual acquisition robot, and acquire visual information; wherein the process of acquiring the visual information is to rotate and collect images at each offset point, and the collection direction rotates 360 degrees in the horizontal plane. A creation execution unit is configured to create a digital twin model based on the visual information and the map of the travel and tourism scene. The calculation process of the offset amount is: In the formula, Indicates the offset. The preset correction factor. This represents the number of all entities along the vertical centerline. Indicates the first The subject level of each subject, Indicates the first The distance between a subject and a midpoint, wherein the distance has a sign, the sign being determined by the directional relationship between the subject and the midpoint, and the direction corresponding to the positive sign is a preset value.

Images