A method and system for building indoor virtual scenes based on Unreal Engine

Abstract

The present invention relates to a method and system for building an indoor virtual scene based on Unreal Engine, and relates to the field of virtual reality technology, including: obtaining a user browsing direction, predicting the probability of interactive switching of sub-virtual scenes, and pre-calling multiple virtual resources in multiple sub-virtual scenes to pre-build virtual scenes according to multiple sub-scene probabilities; predicting multiple sub-scene browsing times, and correcting multiple sub-scene browsing times according to multiple sub-scene probabilities; obtaining multiple special effect accuracies according to multiple corrected sub-scene browsing time decisions, pre-loading special effects for multiple sub-virtual scenes, and completing the construction of indoor virtual scenes. The present invention can solve the technical problems in the existing virtual scene construction method that the real-time and accuracy of resource loading are insufficient, resulting in low operating efficiency and inability to meet real-time interaction requirements, and can effectively improve the real-time and accuracy of resource loading, thereby improving the overall operating efficiency and ensuring a smooth real-time interactive experience.

Description

Technical Field

[0001] The present invention relates to the field of virtual reality technology, and in particular to a method and system for building an indoor virtual scene based on an Unreal Engine. Background Art

[0002] With the rapid development of virtual reality (VR) and augmented reality (AR) technologies, indoor virtual scene construction is becoming more and more widely used, especially in the fields of virtual decoration, house viewing, interior design, etc., providing a highly immersive and interactive experience. Through virtual reality technology, users can browse, interact and customize virtual space in an immersive way.

[0003] In the traditional method of building indoor virtual scenes, the virtual environment usually needs to load a large amount of resources at one time when the user enters the scene, including static models, textures, dynamic effects, and special effects. Although this full loading method can provide a complete scene, the huge amount of resources required during the loading process leads to insufficient real-time and accuracy of resource loading, low overall operating efficiency, and failure to meet the user's real-time interaction needs in the virtual scene. Summary of the invention

[0004] The present invention aims to solve the technical problems that the existing indoor virtual scene construction method has low operating efficiency and cannot meet the real-time interaction requirements due to insufficient real-time and accuracy of resource loading, and provides an indoor virtual scene construction method and system based on Unreal Engine to solve the problem.

[0005] The technical solution of the present invention to solve the above technical problems is as follows:

[0006] In a first aspect, the present invention provides a method for building an indoor virtual scene based on an Unreal Engine, comprising: in response to a user's virtual scene browsing instruction, building a main virtual scene in an indoor virtual scene based on the Unreal Engine; obtaining the user's browsing direction in the main virtual scene, predicting the probability of interactive switching of other multiple sub-virtual scenes, obtaining multiple sub-scene probabilities, and pre-calling multiple virtual resources in the multiple sub-virtual scenes to pre-build the virtual scene according to the multiple sub-scene probabilities; predicting the user's browsing time of multiple sub-scenes in the multiple sub-virtual scenes according to the user's historical browsing data, and performing auxiliary correction on the multiple sub-scene browsing time according to the multiple sub-scene probabilities to obtain multiple corrected sub-scene browsing times; deciding to obtain multiple special effect accuracies according to the multiple corrected sub-scene browsing times, preloading special effects for the pre-built multiple sub-virtual scenes, and completing the construction of the indoor virtual scene.

[0007] In a second aspect, the present invention provides an indoor virtual scene construction system based on Unreal Engine, comprising: a main virtual scene construction module, which is used to respond to the user's virtual scene browsing instructions and, based on Unreal Engine, build a main virtual scene in the indoor virtual scene; a sub-virtual scene pre-construction module, which is used to obtain the user's browsing direction in the main virtual scene, predict the interactive switching probability of other multiple sub-virtual scenes, obtain multiple sub-scene probabilities, and pre-call multiple virtual resources in the multiple sub-virtual scenes to pre-construct the virtual scene according to the multiple sub-scene probabilities; a sub-scene browsing time correction module, which is used to predict the user's multiple sub-scene browsing time in the multiple sub-virtual scenes according to the user's historical browsing data, and perform auxiliary correction on the multiple sub-scene browsing time according to the multiple sub-scene probabilities to obtain multiple corrected sub-scene browsing times; a sub-virtual scene loading module, which is used to decide to obtain multiple special effect accuracies according to the multiple corrected sub-scene browsing times, pre-load special effects for the pre-constructed multiple sub-virtual scenes, and complete the indoor virtual scene construction.

[0008] In a third aspect, the present invention further provides an electronic device, comprising:

[0009] At least one processor; a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor so that the at least one processor can perform the steps of any one of the methods described in the first aspect above.

[0010] In a fourth aspect, a computer-readable storage medium is provided, wherein a computer program is stored on the computer-readable storage medium, and when the computer program is executed, the steps of the method described in any one of the first aspects are implemented.

[0011] The beneficial effects of the present invention are as follows: by responding to the user's virtual scene browsing instructions, based on the Unreal Engine, a main virtual scene in an indoor virtual scene is built; then the browsing direction of the user in the main virtual scene is obtained, and the interactive switching probability of other multiple sub-virtual scenes is predicted to obtain multiple sub-scene probabilities, and multiple virtual resources in the multiple sub-virtual scenes are pre-called according to the multiple sub-scene probabilities to pre-build the virtual scene; further, according to the user's historical browsing data, the browsing time of multiple sub-scenes of the user in the multiple sub-virtual scenes is predicted, and according to the multiple sub-scene probabilities, the browsing time of the multiple sub-scenes is auxiliary corrected to obtain multiple corrected sub-scene browsing times; finally, according to the multiple corrected sub-scene browsing times, a decision is made to obtain multiple special effect accuracies, and special effects are preloaded for the pre-built multiple sub-virtual scenes to complete the construction of the indoor virtual scene; that is, by integrating componentized design, user behavior prediction and dynamic loading optimization technology, the real-time and accuracy of resource loading can be effectively improved, and the switching loading response efficiency of the scene can be significantly improved, thereby improving the overall operation efficiency and ensuring a smooth real-time interactive experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 A schematic diagram of the process of building an indoor virtual scene based on Unreal Engine provided by the present invention;

[0013] Figure 2 A schematic diagram of the structure of an indoor virtual scene construction system based on Unreal Engine provided by the present invention;

[0014] Figure 3 A schematic diagram of the structure of an electronic device provided by the present invention;

[0015] Figure 4 A schematic diagram of the structure of a computer-readable storage medium provided by the present invention.

[0016] In the accompanying drawings, the components represented by the reference numerals are described as follows:

[0017] Main virtual scene building module 01, sub-virtual scene pre-building module 02, sub-scene browsing time correction module 03, sub-virtual scene loading module 04, electronic device 500, memory 510, processor 520, first computer program 511, computer readable storage medium 600, second computer program 611. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0019] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0020] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or explanation". Any embodiment described as "for example" in the present invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any technician in the field to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes will not be elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed in the present invention.

[0021] Embodiment 1, as Figure 1 As shown, an embodiment of the present invention provides a method for building an indoor virtual scene based on Unreal Engine, which specifically includes the following steps:

[0022] S100: In response to a virtual scene browsing instruction from a user, a main virtual scene in an indoor virtual scene is constructed based on the Unreal Engine.

[0023] Furthermore, step S100 of the present invention further includes:

[0024] S110: In response to the user's virtual scene browsing instruction, the main scene resources and main scene special effects of the main virtual scene in the indoor virtual scene are retrieved; S120: Based on the Unreal Engine, the main virtual scene is constructed and rendered according to the main scene resources and main scene special effects; S130: An interactive button is constructed in the main virtual scene for interactively switching to multiple sub-virtual scenes in the indoor virtual scene to complete the construction of the main virtual scene.

[0025] Specifically, first, the virtual scene browsing instructions issued by the user are identified through the controller of the VR device or user input (such as clicking a button, voice commands, etc.). For example, when viewing interior decoration through VR, when the user chooses to view the indoor virtual scene, the main virtual scene (such as the living room) will be loaded and rendered according to the instructions; then, according to the browsing instructions, the main scene resources and main scene special effects of the main virtual scene in the indoor virtual scene are called up. For example, the resources required for the main virtual scene are called up from the resource library, including static elements (walls, floors, ceilings, etc.), dynamic elements (furniture, lights, home appliances, etc.), and special effects resources (such as lighting, shadows, reflections, etc.). The types of special effects can be dynamically configured according to user needs. For example, the living room scene may require fine lighting adjustment and high-precision light and shadow effects.

[0026] Unreal Engine can provide powerful graphics rendering capabilities, support high-precision lighting, shadows, reflections, material processing and physical simulation, and can create realistic and delicate 3D scenes, which is especially suitable for application scenarios that require high-fidelity graphics rendering; then, using the powerful modeling and rendering capabilities of Unreal Engine, load the retrieved main scene resources and main scene special effects, and display them in a virtual reality environment to build the main virtual scene. For example, by utilizing Unreal Engine's lighting, materials, shadows and other rendering technologies, the realism and immersion of the main virtual scene can be enhanced.

[0027] Then, in the main virtual scene, interactive buttons for interactively switching to multiple sub-virtual scenes in the indoor virtual scene are constructed. For example, multiple interactive buttons (such as "enter the master bedroom", "enter the kitchen", etc.) are designed. These buttons allow users to switch to the corresponding sub-virtual scenes according to their needs. Each button corresponds to the entrance of a sub-scene and triggers a specific switching behavior when the user clicks it. Among them, interactive buttons can be created through the UI design tool of Unreal Engine (such as UMG). These buttons can be triggered by different interactive methods (such as clicking, floating, dragging, etc.). When the user clicks an interactive button, the system will load the corresponding sub-scene resources according to the currently browsed scene and the user's needs. For example, when clicking the "enter the kitchen" button, the virtual scene will dynamically load the resources and special effects of the kitchen. The construction of the main virtual scene is completed through the above steps. At this time, the user can see the complete main virtual scene (such as the living room) and various elements in the scene, such as furniture, lights, curtains, etc. At the same time, the interactive buttons of multiple sub-virtual scenes have been loaded and can be used by the user.

[0028] By loading the main virtual scene and its resources and special effects on demand, we can avoid loading too much content at one time, reduce the system burden and improve the loading speed; through dynamic loading and interactive button design, the system can quickly respond to user operations, allowing users to browse and switch different virtual scenes smoothly.

[0029] S200: Obtain the browsing direction of the user in the main virtual scene, predict the interactive switching probability of other multiple sub-virtual scenes, obtain multiple sub-scene probabilities, and pre-call multiple virtual resources in the multiple sub-virtual scenes to pre-build the virtual scene based on the multiple sub-scene probabilities.

[0030] Furthermore, step S200 of the present invention further includes:

[0031] S210: Obtaining the browsing direction of the user in the main virtual scene.

[0032] Specifically, first, the browsing direction of the user in the main virtual scene is obtained. For example, the user's viewing angle is 45° east of south, that is, the user is currently facing the southeast direction of the main virtual scene. According to the user's direction, the sub-scene that the user is more likely to enter can be inferred. For example, assuming that the kitchen is located in the southeast direction and the bedroom is located in the northwest direction, then when the user faces southeast, the probability of entering the kitchen is higher.

[0033] S220: Pre-training a switching probability prediction path for predicting the interactive switching probabilities of the multiple sub-virtual scenes.

[0034] Further, step S220 of the present invention further includes:

[0035] S221: Based on the browsing records of the indoor virtual scene in the historical time, a set of sample browsing directions is collected, and the proportion of user interaction switching to the multiple sub-virtual scenes after different sample browsing directions is obtained, which is marked as a sample sub-scene probability distribution set, wherein each sample sub-scene probability distribution includes multiple sample sub-scene probabilities of switching to multiple sub-virtual scenes; S222: The sample browsing direction set and the sample sub-scene probability distribution set are used as supervised training data and test data to perform supervised training and testing of the switching probability prediction path; S223: After the test meets the requirements, the constructed switching probability prediction path is obtained.

[0036] Specifically, first, the browsing records of the indoor virtual scene in the historical time (such as within the last month) are obtained, and sample browsing directions are collected according to the browsing records. By collecting enough browsing direction data, a direction set containing multiple samples is formed to obtain a sample browsing direction set. Further obtain the proportion of the user interactively switching to the multiple sub-virtual scenes after different sample browsing directions, and obtain the switching proportion of the sub-virtual scenes by calculating the average value through multiple sample data. For example, assuming that the user visits the following sub-scenes 20 times, including the master bedroom 40 times, the kitchen 40 times, the living room 30 times, and the study 10 times when the user browses in the browsing direction of "east by south 45°", the total number of switching times is 100 times, the master bedroom switching proportion is 20%, the kitchen switching proportion is 40%, the living room switching proportion is 30%, and the study switching proportion is 10% when the user browses in the browsing direction of "east by south 45°". The switching proportion of the sub-virtual scenes is identified as the sample sub-scene probability, and the sample sub-scene probability distribution after different sample browsing directions is obtained, wherein each sample sub-scene probability distribution includes multiple sample sub-scene probabilities of switching to multiple sub-virtual scenes, and a sample sub-scene probability distribution set is obtained.

[0037] A switching probability prediction path is constructed based on a BP neural network. The switching probability prediction path is a BP neural network model that can be iteratively optimized in machine learning, including an input layer, multiple hidden layers and an output layer. The input data of the input layer is the browsing direction, the output data of the output layer is the sub-scene probability distribution, and the hidden layer is used to perform multi-level nonlinear transformation on the input data to extract deeper features. Through the multi-level hidden layers, the network can capture the relationship between complex browsing patterns and sub-scene switching. Then, the sample browsing direction set and the sample sub-scene probability distribution set are used as sample data sets, and the sample data sets are divided into a training set and a verification set according to a predetermined ratio (such as 85% for the training set and 15% for the verification set); then, the sample browsing direction is used as input, and the sample sub-scene probability distribution is used as supervision, and the training set and the verification set are used to perform supervised training and testing on the switching probability prediction path. First, the browsing direction of each sample in the training set is input into the neural network, and the predicted sub-scene switching probability is calculated; then, by comparing with the actual sub-scene probability distribution, the loss function is calculated, and the weights and biases of the neural network are adjusted by the back propagation algorithm, and the loss function is minimized by the gradient descent method. The training process will go through multiple iterations, and each iteration network will update the weights according to the error and gradually optimize the model; when the predetermined number of training times is met, the switching probability prediction path is tested by the verification set until the predetermined test requirements are met (such as the output accuracy is greater than 90%), and a trained switching probability prediction path is obtained.

[0038] By constructing a switching probability prediction path based on the BP neural network to predict the interactive switching probability of multiple sub-virtual scenes, the accuracy and efficiency of the sub-virtual scene switching probability prediction can be improved, so that user behavior can be accurately predicted, providing a reliable basis for the next step of sub-virtual scene resource loading, and ensuring the rationality of resource loading priority.

[0039] S230: Input the browsing direction into the switching probability prediction path, and predict and output to obtain multiple sub-scene probabilities; S240: Configure multiple sub-scene resource call ratios according to the multiple sub-scene probabilities; S250: According to the multiple sub-scene resource call ratios, call the sub-scene resources in the multiple sub-virtual scenes to obtain multiple virtual resources; S260: Pre-build the multiple sub-virtual scenes according to the multiple virtual resources.

[0040] Specifically, the browsing direction is input into the switching probability prediction path for prediction, and multiple sub-scene probabilities are output; multiple sub-scene resource retrieval ratios are further configured according to the multiple sub-scene probabilities, and matching resource retrieval ratios are set according to the sub-scene probabilities. For example, assuming that the kitchen probability is 70%, the master bedroom probability is 20%, and the guest bedroom probability is 10%, the resource retrieval ratio is set according to the corresponding probability, that is, the kitchen resource retrieval ratio is 70%, the master bedroom resource retrieval ratio is 20%, and the guest bedroom resource retrieval ratio is 10%. Then, according to the multiple sub-scene resource retrieval ratios, the sub-scene resources in the multiple sub-virtual scenes are retrieved, and the sub-scene resources include but are not limited to static resources (such as scene models, furniture, decorations, etc.), dynamic resources (such as lighting effects, particle special effects, etc.), and interactive resources (such as buttons, switching interfaces, sounds, etc.), to obtain multiple virtual resources. Exemplarily, resources with higher importance are retrieved according to the resource retrieval ratio. The multiple types of resources in each sub-virtual scene may be sorted in advance (for example, the model resources of the bed in the bedroom are sorted first, and the model resources of the wardrobe are sorted last). Then, resources with the top resource retrieval ratio are retrieved, for example, the top 70% of the resources are retrieved. Multiple resources may be retrieved by multiplying the number of resources by 70% and then rounding up.

[0041] Finally, the multiple sub-virtual scenes are pre-built (model rendering, interaction settings, etc.) according to the multiple virtual resources. The pre-built sub-scenes enable users to see smooth effects when switching to the sub-scenes without waiting for resources to load, which improves the response speed of scene switching. At the same time, due to the predicted resource retrieval ratio, the system only loads the sub-scene resources that users are most likely to switch to, thereby saving system resources and improving efficiency. Through the above technical process, the switching of sub-virtual scenes and resource loading can be more intelligent, accurate and efficient, greatly improving the performance of the overall system and user experience.

[0042] S300: Predicting browsing times of multiple sub-scenes of the user in the multiple sub-virtual scenes based on the historical browsing data of the user, and performing auxiliary corrections on the multiple sub-scene browsing times based on the multiple sub-scene probabilities to obtain multiple corrected sub-scene browsing times.

[0043] Furthermore, step S300 of the present invention further includes:

[0044] S310: Based on the historical browsing data of the user, a sample indoor scene feature information set of multiple sample indoor virtual scenes is collected, and the browsing time of the user in different sub-virtual scenes in the multiple sample indoor virtual scenes is obtained to obtain multiple sample sub-scene browsing time sets; S320: Using the sample indoor scene feature information set and the multiple sample sub-scene browsing time sets as supervised training data and test data, training the browsing time prediction path and testing it until the test meets the requirements; S330: Obtain the indoor scene feature information of the indoor virtual scene, input the browsing time prediction path, and predict and output to obtain multiple sub-scene browsing times.

[0045] Specifically, based on the historical browsing data of the user, a set of sample indoor scene feature information of multiple sample indoor virtual scenes is collected, wherein the indoor scene feature information includes apartment type (such as one-bedroom, two-bedroom, three-bedroom, etc.), area, orientation (such as east, south, west, north), furniture arrangement, etc., wherein the spatial layout of different apartment types will affect the user's interest in and stay time in certain scenes; a large room may cause the user to stay in the area for a long time; the orientation has an impact on indoor lighting, air circulation, etc., which may indirectly affect the user's preference for the area, etc. The browsing time of the user in different sub-virtual scenes in the multiple sample indoor virtual scenes is further obtained, such as the browsing time of the kitchen is 3 minutes, the browsing time of the living room is 5 minutes, etc., and multiple sample sub-scene browsing time sets are obtained.

[0046] Then, a browsing time prediction path is constructed based on the BP neural network, and the browsing time prediction path includes an input layer, multiple hidden layers and an input layer, wherein the input data of the input layer is indoor scene feature information, including parameters such as apartment type, area, and orientation; the output data of the output is multiple sub-scene browsing time, that is, the time the user may stay in each sub-virtual scene, for example, the predicted browsing time of the user in the kitchen is 4 minutes, the living room is 3 minutes, the bedroom is 2 minutes, etc.; the hidden layer is used to capture nonlinear relationships using multiple hidden layers, and gradually map complex data patterns to prediction results. Each hidden layer contains a number of neurons, which process the input data and extract features through activation functions (such as ReLU). The output of each layer is used as the input of the next layer, and more abstract features are gradually extracted to predict the browsing time of users in different sub-scenes.

[0047] Then, the sample indoor scene feature information set and the browsing time sets of multiple sample sub-scenes are used as sample data sets, and the sample data sets are divided into a training set and a verification set according to a predetermined ratio (such as 85% for training set and 15% for verification set); further, the sample indoor scene feature information is used as input, and the browsing time of multiple sample sub-scenes is used as supervision, and the browsing time prediction path is supervised and trained using the training set. First, the input indoor scene feature information is passed to the hidden layer of the network using the forward propagation step of the BP neural network, and a nonlinear transformation is performed through an activation function (such as ReLU), and finally the browsing time prediction of each sub-scene is output; then, the error between the output prediction result and the actual browsing time value is calculated through a loss function such as the mean square error; then, the error is transmitted from the output layer back to the input layer through a back propagation algorithm, and the weights and biases in the network are updated to minimize the error; through an iterative method, the network will continuously adjust parameters to optimize its prediction accuracy. At the same time, after each training cycle, the trained model is tested using a validation set. The validation set contains data that is different from the training set, which can help evaluate the generalization ability of the model. When the output result is greater than the predetermined test accuracy index, the browsing time prediction path of the training is obtained.

[0048] Then, indoor scene feature information of the indoor virtual scene is obtained, such as a two-bedroom apartment, an area of ​​96 square meters, and a south orientation, etc.; and the indoor scene feature information is input into the browsing time prediction path for prediction, and multiple sub-scene browsing times are output.

[0049] S340: performing auxiliary correction on the multiple sub-scene browsing times according to the multiple sub-scene probabilities to obtain multiple corrected sub-scene browsing times.

[0050] Further, step S340 of the present invention further includes:

[0051] S341: Calculate the mean of the multiple sub-scene probabilities to obtain the average probability; S342: Calculate the ratio of each sub-scene probability to the average probability to obtain multiple browsing time correction coefficients; S343: Multiply the multiple sub-scene browsing times by the multiple browsing time correction coefficients respectively, perform auxiliary correction calculations, and obtain multiple corrected sub-scene browsing times.

[0052] Specifically, the mean of the multiple sub-scene probabilities is calculated to obtain an average probability; then the ratio of each sub-scene probability to the average probability is calculated respectively, and the ratio is set as a browsing time correction coefficient to obtain multiple browsing time correction coefficients, wherein the correction coefficient represents the deviation of the switching probability of each sub-scene relative to the average probability, and reflects the user's interest in and expected browsing time of a particular sub-scene. For example, if the user is more interested in the kitchen, then the switching probability of the kitchen will be higher than other scenes. At this time, the correction coefficient of the kitchen will be greater than 1, indicating that the time the user stays in the kitchen scene should be appropriately increased; if the user is less interested in the bedroom, the switching probability of the bedroom will be smaller, and the correction coefficient will be less than 1, thereby reducing the browsing time of the bedroom.

[0053] Then, the browsing time correction coefficients are multiplied by the browsing time of the sub-scenes, that is, the browsing time correction coefficient is multiplied by the predicted browsing time of the same sub-scene to perform auxiliary correction calculation. For example, assuming that the browsing time correction coefficient of the living room is 0.8 and the predicted browsing time of the living room is 5 minutes, the corrected browsing time of the living room is 0.8. 5=4 minutes; multiple corrected sub-scene browsing times are calculated. By calculating the correction coefficient to correct the sub-scene browsing time, the accuracy and reliability of the sub-scene browsing time can be improved, so that the computing and rendering resources can be allocated more reasonably, avoiding excessive resources being wasted on low-probability and low-interest sub-scenes, thereby improving the overall operating efficiency, significantly improving the interactivity and fluency of the virtual scene, and allowing users to have a more realistic and comfortable experience during the browsing process.

[0054] S400: According to the multiple corrected sub-scene browsing times, a decision is made to obtain multiple special effect precisions, and special effects are preloaded for the pre-built multiple sub-virtual scenes to complete the construction of the indoor virtual scene.

[0055] Furthermore, step S400 of the present invention further includes:

[0056] S410: Obtain multiple sample browsing time intervals and multiple sample special effect accuracy levels; S420: Use the multiple sample browsing time intervals and multiple sample special effect accuracy levels to construct a special effect accuracy decision path based on a decision tree; S430: Input the multiple corrected sub-scene browsing times into the special effect accuracy decision path respectively, and obtain multiple special effect accuracies through decision classification; S440: Preload special effects in the pre-constructed multiple sub-virtual scenes according to the multiple special effect accuracies respectively.

[0057] Specifically, first, based on the browsing records of the indoor virtual scene in the historical time, multiple sample browsing time intervals are obtained, that is, the browsing time range of users for different sub-scenes, such as short browsing (browsing time less than 2 minutes), medium browsing (browsing time 2 to 5 minutes), and long browsing (browsing time greater than 5 minutes); further, the sample special effect accuracy levels (such as lighting effect, shadow effect, reflection accuracy, etc.) corresponding to different sample browsing time intervals are obtained, wherein the longer the browsing time, the higher the corresponding special effect accuracy level, wherein the special effect accuracy levels include low accuracy (low image quality, low lighting efficiency, low shadow quality), medium accuracy (medium image quality, moderate lighting effects and shadows), high accuracy (high image quality, real lighting effects, advanced shadows, etc.), etc., which can be limited by specific numerical indicators to obtain multiple sample special effect accuracy levels.

[0058] Then, based on the decision tree principle, the sample browsing time interval is used as the child node of the tree, the corresponding sample special effect accuracy level is used as the leaf node of the child node, and the multiple sample browsing time intervals and multiple sample special effect accuracy levels are used as construction data to build a special effect accuracy decision path. Through the special effect accuracy decision path based on the decision tree, the special effect accuracy of each sub-scene can be accurately classified according to the user's browsing time interval.

[0059] The multiple corrected sub-scene browsing times are further input into the special effect precision decision path for decision classification, and multiple special effect precisions are output. Finally, the special effects in the pre-built multiple sub-virtual scenes are preloaded according to the multiple special effect precisions to complete the indoor virtual scene construction. Through the above steps, the special effects of all sub-scenes have adjusted their precision according to the user's browsing behavior and interests, and finally the indoor virtual scene is built. This ensures that while saving computing resources, it can also provide users with a high-quality scene experience, ensure that the loading response efficiency of the virtual scene is balanced with the picture quality, and significantly improve the user experience.

[0060] The method for building an indoor virtual scene based on Unreal Engine provided by the embodiment of the present invention has at least the following technical effects:

[0061] In response to the user's virtual scene browsing instructions, a main virtual scene in the indoor virtual scene is built based on the Unreal Engine; then the user's browsing direction in the main virtual scene is obtained, and the interactive switching probability of other multiple sub-virtual scenes is predicted to obtain multiple sub-scene probabilities, and multiple virtual resources in the multiple sub-virtual scenes are pre-called according to the multiple sub-scene probabilities to pre-build the virtual scene; further, according to the user's historical browsing data, the user's multiple sub-scene browsing time in the multiple sub-virtual scenes is predicted, and according to the multiple sub-scene probabilities, the multiple sub-scene browsing time is auxiliary corrected to obtain multiple corrected sub-scene browsing time; finally, according to the multiple corrected sub-scene browsing time, a decision is made to obtain multiple special effect accuracies, and special effects are preloaded for the pre-built multiple sub-virtual scenes to complete the indoor virtual scene construction; that is, by integrating componentized design, user behavior prediction and dynamic loading optimization technology, the real-time and accuracy of resource loading can be effectively improved, and the switching loading response efficiency of the scene can be significantly improved, thereby improving the overall operation efficiency and ensuring a smooth real-time interactive experience.

[0062] Embodiment 2, as Figure 2 As shown, based on the same inventive concept as the method for building an indoor virtual scene based on the Unreal Engine provided in the first embodiment, the embodiment of the present invention further provides a system for building an indoor virtual scene based on the Unreal Engine, comprising:

[0063] The main virtual scene construction module 01 is used to respond to the user's virtual scene browsing instruction and build the main virtual scene in the indoor virtual scene based on the Unreal Engine; the sub-virtual scene pre-construction module 02 is used to obtain the user's browsing direction in the main virtual scene, predict the interactive switching probability of other multiple sub-virtual scenes, obtain multiple sub-scene probabilities, and pre-call multiple virtual resources in the multiple sub-virtual scenes to pre-construct the virtual scene according to the multiple sub-scene probabilities; the sub-scene browsing time correction module 03 is used to predict the user's multiple sub-scene browsing time in the multiple sub-virtual scenes according to the user's historical browsing data, and perform auxiliary correction on the multiple sub-scene browsing time according to the multiple sub-scene probabilities to obtain multiple corrected sub-scene browsing times; the sub-virtual scene loading module 04 is used to decide to obtain multiple special effect accuracies according to the multiple corrected sub-scene browsing times, pre-load special effects for the pre-constructed multiple sub-virtual scenes, and complete the indoor virtual scene construction.

[0064] Furthermore, the indoor virtual scene construction system based on the Unreal Engine also includes: in response to the user's virtual scene browsing instructions, calling up the main scene resources and main scene special effects of the main virtual scene in the indoor virtual scene; based on the Unreal Engine, constructing and rendering the main virtual scene according to the main scene resources and main scene special effects; constructing an interactive button in the main virtual scene for interactively switching to multiple sub-virtual scenes in the indoor virtual scene to complete the construction of the main virtual scene.

[0065] Furthermore, the indoor virtual scene construction system based on the Unreal Engine also includes: obtaining the browsing direction of the user in the main virtual scene; pre-training a switching probability prediction path for predicting the interactive switching probability of the multiple sub-virtual scenes; inputting the browsing direction into the switching probability prediction path, and predicting and outputting multiple sub-scene probabilities; configuring multiple sub-scene resource retrieval ratios according to the multiple sub-scene probabilities; according to the multiple sub-scene resource retrieval ratios, retrieving sub-scene resources in the multiple sub-virtual scenes to obtain multiple virtual resources; and pre-building the multiple sub-virtual scenes according to the multiple virtual resources.

[0066] Furthermore, the indoor virtual scene construction system based on the Unreal Engine also includes: collecting a set of sample browsing directions according to the browsing records of the indoor virtual scenes in historical time, and obtaining the proportion of users interactively switching to the multiple sub-virtual scenes after different sample browsing directions, and marking them as a sample sub-scene probability distribution set, wherein each sample sub-scene probability distribution includes multiple sample sub-scene probabilities of switching to multiple sub-virtual scenes; using the sample browsing direction set and the sample sub-scene probability distribution set as supervised training data and test data to perform supervised training and testing of the switching probability prediction path; after the test meets the requirements, obtaining the constructed switching probability prediction path.

[0067] Furthermore, the indoor virtual scene construction system based on the Unreal Engine also includes: collecting a sample indoor scene feature information set of multiple sample indoor virtual scenes according to the user's historical browsing data, and obtaining the user's browsing time in different sub-virtual scenes in the multiple sample indoor virtual scenes to obtain multiple sample sub-scene browsing time sets; using the sample indoor scene feature information set and the multiple sample sub-scene browsing time sets as supervised training data and test data, training the browsing time prediction path and testing until the test meets the requirements; obtaining the indoor scene feature information of the indoor virtual scene, inputting the browsing time prediction path, and predicting and outputting multiple sub-scene browsing times; and performing auxiliary corrections on the multiple sub-scene browsing times according to the multiple sub-scene probabilities to obtain multiple corrected sub-scene browsing times.

[0068] Furthermore, the indoor virtual scene construction system based on Unreal Engine also includes: calculating the mean of the probabilities of the multiple sub-scenes to obtain an average probability; calculating the ratio of each sub-scene probability to the average probability to obtain multiple browsing time correction coefficients; and multiplying the multiple sub-scene browsing times by the multiple browsing time correction coefficients to perform auxiliary correction calculations to obtain multiple corrected sub-scene browsing times.

[0069] Furthermore, the indoor virtual scene construction system based on the Unreal Engine also includes: obtaining multiple sample browsing time intervals and multiple sample special effect accuracy levels; using the multiple sample browsing time intervals and multiple sample special effect accuracy levels, based on a decision tree, constructing a special effect accuracy decision path; inputting the multiple corrected sub-scene browsing times into the special effect accuracy decision path, and obtaining multiple special effect accuracies through decision classification; and preloading special effects in multiple pre-constructed sub-virtual scenes according to the multiple special effect accuracies.

[0070] For example 3, please refer to Figure 3 , Figure 3 Schematic diagram of an electronic device provided by an embodiment of the present invention. Figure 3 As shown, an embodiment of the present invention provides an electronic device 500, including a memory 510, a processor 520, and a first computer program 511 stored in the memory 510 and executable on the processor 520. When the processor 520 executes the first computer program 511, the following steps are implemented: in response to a user's virtual scene browsing instruction, based on the Unreal Engine, a main virtual scene in an indoor virtual scene is constructed; the browsing direction of the user in the main virtual scene is obtained, and the interactive switching probability of other multiple sub-virtual scenes is predicted to obtain multiple sub-scene probabilities, and multiple virtual resources in the multiple sub-virtual scenes are pre-retrieved according to the multiple sub-scene probabilities to pre-construct the virtual scene; according to the user's historical browsing data, the browsing time of multiple sub-scenes of the user in the multiple sub-virtual scenes is predicted, and according to the multiple sub-scene probabilities, the browsing time of the multiple sub-scenes is auxiliary corrected to obtain multiple corrected sub-scene browsing times; according to the multiple corrected sub-scene browsing times, a decision is made to obtain multiple special effect accuracies, and special effects are preloaded for the pre-constructed multiple sub-virtual scenes to complete the construction of the indoor virtual scene.

[0071] Example 4, please refer to Figure 4 , Figure 4 A schematic diagram of an embodiment of a computer-readable storage medium provided in an embodiment of the present invention. Figure 4As shown, this embodiment provides a computer-readable storage medium 600, on which a second computer program 611 is stored. When the second computer program 611 is executed by the processor, the following steps are implemented: in response to the user's virtual scene browsing instruction, based on the Unreal Engine, a main virtual scene in the indoor virtual scene is built; the browsing direction of the user in the main virtual scene is obtained, and the interactive switching probability of other multiple sub-virtual scenes is predicted to obtain multiple sub-scene probabilities, and multiple virtual resources in the multiple sub-virtual scenes are pre-called according to the multiple sub-scene probabilities to pre-build the virtual scene; according to the user's historical browsing data, the browsing time of multiple sub-scenes of the user in the multiple sub-virtual scenes is predicted, and according to the multiple sub-scene probabilities, the browsing time of the multiple sub-scenes is auxiliary corrected to obtain multiple corrected sub-scene browsing times; according to the multiple corrected sub-scene browsing times, a decision is made to obtain multiple special effect accuracies, and special effects are preloaded for the pre-built multiple sub-virtual scenes to complete the construction of the indoor virtual scene.

[0072] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0073] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0074] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0075] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0076] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0077] Although preferred embodiments of the present invention have been described, additional changes and modifications may occur to these embodiments once those skilled in the art understand the basic inventive concepts.

[0078] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention belong to the scope of the present invention and its equivalent technologies, the present invention is also intended to include these changes and variations.

Claims

1. A method for building an indoor virtual scene based on Unreal Engine, characterized in that: Methods include: In response to the user's virtual scene browsing instruction, a main virtual scene in the indoor virtual scene is built based on the Unreal Engine; Obtaining the browsing direction of the user in the main virtual scene, predicting the probability of interactive switching of multiple other sub-virtual scenes, obtaining multiple sub-scene probabilities, and pre-calling multiple virtual resources in the multiple sub-virtual scenes to pre-build the virtual scene according to the multiple sub-scene probabilities; Predicting browsing times of multiple sub-scenes of the user in the multiple sub-virtual scenes according to the historical browsing data of the user, and performing auxiliary corrections on the browsing times of the multiple sub-scenes according to the multiple sub-scene probabilities to obtain multiple corrected browsing times of the sub-scenes; According to the multiple corrected sub-scene browsing times, a decision is made to obtain multiple special effect precisions, and special effects are preloaded for the pre-built multiple sub-virtual scenes to complete the indoor virtual scene construction; The method includes obtaining the browsing direction of the user in the main virtual scene, predicting the probability of interactive switching of multiple sub-virtual scenes, obtaining multiple sub-scene probabilities, and pre-calling multiple virtual resources in the multiple sub-virtual scenes to pre-build the virtual scene according to the multiple sub-scene probabilities, including: Obtaining the browsing direction of the user in the main virtual scene; Pre-training a switching probability prediction path, wherein the switching probability prediction path is used to perform interactive switching probability prediction of the multiple sub-virtual scenes; Inputting the browsing direction into the switching probability prediction path, and predicting and outputting a plurality of sub-scene probabilities; According to the multiple sub-scenario probabilities, configuring multiple sub-scenario resource call ratios; Retrieving sub-scene resources in the multiple sub-virtual scenes according to the resource retrieval ratio of the multiple sub-scenes to obtain multiple virtual resources; Pre-constructing the multiple sub-virtual scenes according to the multiple virtual resources; Among them, the pre-trained switching probability prediction path includes: According to the browsing records of the indoor virtual scene in the historical time, a set of sample browsing directions is collected, and the proportions of users interactively switching to the multiple sub-virtual scenes after different sample browsing directions are obtained, which are marked as a sample sub-scene probability distribution set, wherein each sample sub-scene probability distribution includes multiple sample sub-scene probabilities of switching to the multiple sub-virtual scenes; Using the sample browsing direction set and the sample sub-scene probability distribution set as supervised training data and test data, to perform supervised training and testing of the switching probability prediction path; After the test meets the requirements, the constructed switching probability prediction path is obtained.

2. The method for building an indoor virtual scene based on Unreal Engine according to claim 1, characterized in that: In response to the user's virtual scene browsing instruction, a main virtual scene in the indoor virtual scene is built based on the Unreal Engine, including: In response to a virtual scene browsing instruction of a user, calling up a main scene resource and a main scene special effect of a main virtual scene in the indoor virtual scene; Based on the Unreal Engine, construct and render the main virtual scene according to the main scene resources and main scene special effects; An interactive button for interactively switching to a plurality of sub-virtual scenes in the indoor virtual scene is constructed in the main virtual scene to complete the construction of the main virtual scene.

3. The method for building an indoor virtual scene based on Unreal Engine according to claim 1, characterized in that: Predicting the browsing time of multiple sub-scenes of the user in the multiple sub-virtual scenes according to the historical browsing data of the user, and performing auxiliary correction on the browsing time of the multiple sub-scenes according to the multiple sub-scene probabilities to obtain multiple corrected sub-scene browsing times, including: According to the historical browsing data of the user, a sample indoor scene feature information set of multiple sample indoor virtual scenes is collected, and the browsing time of the user in different sub-virtual scenes in the multiple sample indoor virtual scenes is obtained to obtain a plurality of sample sub-scene browsing time sets; Using the sample indoor scene feature information set and multiple sample sub-scene browsing time sets as supervised training data and test data, training the browsing time prediction path and performing tests until the tests meet the requirements; Acquire sample indoor scene feature information of the sample indoor virtual scene, input the browsing time prediction path, and predict and output to obtain browsing times of multiple sub-scenes; According to the multiple sub-scene probabilities, auxiliary corrections are performed on the multiple sub-scene browsing times to obtain multiple corrected sub-scene browsing times.

4. The method for building an indoor virtual scene based on Unreal Engine according to claim 3, characterized in that: According to the multiple sub-scene probabilities, auxiliary corrections are performed on the multiple sub-scene browsing times to obtain multiple corrected sub-scene browsing times, including: Calculating the average of the multiple sub-scenario probabilities to obtain an average probability; Calculating the ratio of each sub-scenario probability to the average probability to obtain multiple browsing time correction coefficients; The plurality of browsing time correction coefficients are respectively multiplied by the plurality of sub-scene browsing times to perform auxiliary correction calculations to obtain a plurality of corrected sub-scene browsing times.

5. The method for building an indoor virtual scene based on Unreal Engine according to claim 1, characterized in that: According to the multiple corrected sub-scene browsing times, a plurality of special effect precisions are determined and preloaded with special effects for the pre-built multiple sub-virtual scenes, including: Obtain multiple sample browsing time intervals and obtain multiple sample special effect accuracy levels; Using the multiple sample browsing time intervals and the multiple sample special effect accuracy levels, based on a decision tree, construct a special effect accuracy decision path; Inputting the multiple corrected sub-scene browsing times into the special effect precision decision path respectively, and obtaining multiple special effect precisions through decision classification; Special effects in a plurality of pre-constructed sub-virtual scenes are preloaded according to the plurality of special effects precisions respectively.

6. An indoor virtual scene construction system based on Unreal Engine, characterized in that: The steps for implementing the method for building an indoor virtual scene based on Unreal Engine as described in any one of claims 1 to 5 include: A main virtual scene building module, used to respond to a user's virtual scene browsing instruction and build a main virtual scene in an indoor virtual scene based on the Unreal Engine; A sub-virtual scene pre-construction module is used to obtain the browsing direction of the user in the main virtual scene, predict the interactive switching probability of other multiple sub-virtual scenes, obtain multiple sub-scene probabilities, and pre-call multiple virtual resources in the multiple sub-virtual scenes to pre-construct the virtual scene according to the multiple sub-scene probabilities; A sub-scene browsing time correction module is used to predict the browsing time of multiple sub-scenes of the user in the multiple sub-virtual scenes according to the historical browsing data of the user, and to perform auxiliary correction on the browsing time of the multiple sub-scenes according to the multiple sub-scene probabilities to obtain multiple corrected browsing time of the sub-scenes; The sub-virtual scene loading module is used to decide to obtain multiple special effect precisions according to the multiple corrected sub-scene browsing times, preload special effects for the pre-built multiple sub-virtual scenes, and complete the construction of the indoor virtual scene.

7. An electronic device, characterized in that: include: Memory for storing computer software programs; A processor is used to read and execute the computer software program, thereby implementing the steps of the method for building an indoor virtual scene based on the Unreal Engine as described in any one of claims 1 to 5.

8. A non-transitory computer-readable storage medium, characterized in that: The storage medium stores a computer software program, which, when executed by a processor, implements the steps of the method for building an indoor virtual scene based on an Unreal Engine as described in any one of claims 1 to 5.

Images