Virtual scene generation method and device, terminal equipment and storage medium

Abstract

The application discloses a virtual scene generation method and device, terminal equipment and a storage medium. The virtual scene generation method comprises the following steps: acquiring a plurality of virtual traffic indicator component models established in advance; combining the plurality of virtual traffic indicator component models based on a preset virtual traffic indicator model combination rule, and generating a corresponding virtual traffic indicator model. According to the application, each component of the virtual traffic indicator model is decoupled and modeled, and is divided into a plurality of reusable virtual traffic indicator component models. The corresponding virtual traffic indicator model is generated in batches and automatically, the human cost consumed in the virtual traffic indicator model generation process is reduced, and the efficiency of the virtual traffic indicator model generation is improved.

Description

Technical Field

[0001] This application relates to the field of virtual scene generation technology, and in particular to a virtual scene generation method, apparatus, terminal device and storage medium. Background Technology

[0002] The development of autonomous driving is accelerating rapidly, and autonomous driving neural networks are a key technology in this field. The training and validation of autonomous driving neural networks require a large amount of data, which can be acquired primarily through two methods: real-vehicle data collection and virtual scene generation. Real-vehicle data collection is too costly and struggles to capture rare or sporadic scene data, while simulation data generated from virtual scenes can solve this problem.

[0003] The virtual scenes involved in autonomous driving neural networks are related to roads and need to include a large number of traffic signs, traffic lights, and other traffic indicators. Currently, most traffic indicators in virtual scenes are generated manually one by one, which is a costly and inefficient traditional method. Summary of the Invention

[0004] The main purpose of this application is to provide a method, apparatus, terminal device and storage medium for generating virtual scenes, aiming to solve the problems of high cost and low efficiency of traditional traffic sign generation methods.

[0005] To achieve the above objectives, this application provides a virtual scene generation method, the virtual scene generation method comprising:

[0006] Obtain several pre-established virtual traffic sign component models;

[0007] Based on preset virtual traffic sign model combination rules, the various virtual traffic sign component models are combined to generate corresponding virtual traffic sign models.

[0008] Optionally, before the step of obtaining several pre-established virtual traffic sign component models, the method further includes:

[0009] Based on the preset road traffic sign standards, define the virtual traffic sign model combination algorithm and the virtual traffic sign combination paradigm;

[0010] Based on the virtual traffic sign model combination algorithm and the virtual traffic sign combination paradigm, the virtual traffic sign model combination rules are obtained.

[0011] Optionally, the step of combining the several virtual traffic sign component models based on preset virtual traffic sign model combination rules to generate a corresponding virtual traffic sign model includes:

[0012] Based on the virtual traffic sign model combination algorithm and the first type of virtual traffic sign combination paradigm, at least one support model, at least one sign model, and at least one texture model are combined to generate the corresponding first type of virtual traffic sign model.

[0013] Optionally, the step of combining the several virtual traffic sign component models based on preset virtual traffic sign model combination rules to generate a corresponding virtual traffic sign model includes:

[0014] Based on the virtual traffic sign model combination algorithm and the second type of virtual traffic sign combination paradigm, at least one sign model and at least one texture model are combined to generate the corresponding second type of virtual traffic sign model.

[0015] Optionally, the step of combining the several virtual traffic sign component models based on preset virtual traffic sign model combination rules to generate a corresponding virtual traffic sign model includes:

[0016] Based on the virtual traffic sign model combination algorithm and the virtual traffic sign combination paradigm, at least one supporting object model and at least one supported object model are combined to generate a corresponding virtual traffic sign model.

[0017] Optionally, after the step of combining the several virtual traffic sign component models based on preset virtual traffic sign model combination rules to generate a corresponding virtual traffic sign model, the method further includes:

[0018] Acquire road point cloud data and road line data of the target virtual scene;

[0019] The road point cloud data and the road line data are analyzed according to the preset virtual traffic sign model placement rules to determine the sign placement points corresponding to the target virtual scene. The virtual traffic sign model placement rules are obtained based on the preset road traffic sign standard definition.

[0020] Place the virtual traffic sign model at the sign placement point.

[0021] Optionally, before the step of obtaining several pre-established virtual traffic sign component models, the method further includes:

[0022] Obtain the preset requirement configuration file;

[0023] The requirement configuration file is parsed to obtain the virtual traffic sign model requirement parameters corresponding to the target virtual scene;

[0024] Based on the virtual traffic sign model requirement parameters, the steps of obtaining several pre-established virtual traffic sign component models and subsequent steps are executed.

[0025] This application embodiment also proposes a virtual scene generation device, the virtual scene generation device comprising:

[0026] The acquisition module is used to acquire several pre-established virtual traffic indicator component models;

[0027] The generation module is used to combine the several virtual traffic sign component models based on preset virtual traffic sign model combination rules to generate corresponding virtual traffic sign models.

[0028] This application also proposes a terminal device, which includes a memory, a processor, and a virtual scene generation program stored in the memory and executable on the processor. When the virtual scene generation program is executed by the processor, it implements the steps of the virtual scene generation method described above.

[0029] This application also proposes a computer-readable storage medium storing a virtual scene generation program, which, when executed by a processor, implements the steps of the virtual scene generation method described above.

[0030] The virtual scene generation method, apparatus, terminal device, and storage medium proposed in this application acquire several pre-established virtual traffic sign component models; based on preset virtual traffic sign model combination rules, the several virtual traffic sign component models are combined to generate corresponding virtual traffic sign models. Based on this application's solution, the components of the virtual traffic sign model are decoupled and modeled, divided into several reusable virtual traffic sign component models. Based on preset virtual traffic sign model combination rules, these several virtual traffic sign component models can be combined to generate corresponding virtual traffic sign models in batches and automatically, reducing the labor costs consumed in the virtual traffic sign model generation process and improving the efficiency of virtual traffic sign model generation. Attached Figure Description

[0031] Figure 1 This is a schematic diagram of the functional modules of the terminal device to which the virtual scene generation device of this application belongs;

[0032] Figure 2 This is a schematic diagram of the first exemplary embodiment of the virtual scene generation method of this application;

[0033] Figure 3This is a schematic diagram of a virtual scene, including a virtual traffic sign model, involved in the virtual scene generation method of this application.

[0034] Figure 4 This is a schematic diagram of the second exemplary embodiment of the virtual scene generation method of this application;

[0035] Figure 5 This is a schematic diagram of the third exemplary embodiment of the virtual scene generation method of this application;

[0036] Figure 6 This is a schematic diagram of the fourth exemplary embodiment of the virtual scene generation method of this application;

[0037] Figure 7 This is a schematic diagram of the fifth exemplary embodiment of the virtual scene generation method of this application;

[0038] Figure 8 This is a schematic diagram of the sixth exemplary embodiment of the virtual scene generation method of this application;

[0039] Figure 9 This is a schematic diagram of the seventh exemplary embodiment of the virtual scene generation method of this application;

[0040] Figure 10 This is a schematic diagram illustrating the process of generating and placing virtual traffic sign models based on a requirements configuration file, which is involved in the virtual scene generation method of this application.

[0041] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0042] It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit this application.

[0043] The main solution of this application embodiment is: to obtain several pre-established virtual traffic sign component models; and to combine the several virtual traffic sign component models based on preset virtual traffic sign model combination rules to generate corresponding virtual traffic sign models. Based on this application solution, the components of the virtual traffic sign model are decoupled and modeled, divided into several reusable virtual traffic sign component models. Based on preset virtual traffic sign model combination rules, these several virtual traffic sign component models can be combined to generate corresponding virtual traffic sign models in batches and automatically, reducing the manpower cost consumed in the virtual traffic sign model generation process and improving the efficiency of virtual traffic sign model generation.

[0044] Specifically, refer to Figure 1 , Figure 1This diagram illustrates the functional modules of the terminal device to which the virtual scene generation apparatus of this application belongs. The virtual scene generation apparatus can be an independent device capable of generating virtual scenes, and it can be implemented on the terminal device in hardware or software form. The terminal device can be a smart mobile terminal with data processing capabilities, such as a mobile phone or tablet computer, or it can be a fixed terminal device or server with data processing capabilities.

[0045] In this embodiment, the terminal device to which the virtual scene generation device belongs includes at least an output module 110, a processor 120, a memory 130, and a communication module 140.

[0046] The memory 130 stores the operating system and the virtual scene generation program. The virtual scene generation device can acquire several pre-established virtual traffic sign component models; based on preset virtual traffic sign model combination rules, it combines several virtual traffic sign component models, and stores the corresponding virtual traffic sign models and other information in the memory 130. The output module 110 can be a display screen, etc. The communication module 140 can include a WIFI module, a mobile communication module, and a Bluetooth module, etc., and communicates with external devices or servers through the communication module 140.

[0047] When the virtual scene generation program in memory 130 is executed by the processor, it performs the following steps:

[0048] Obtain several pre-established virtual traffic sign component models;

[0049] Based on preset virtual traffic sign model combination rules, the various virtual traffic sign component models are combined to generate corresponding virtual traffic sign models.

[0050] Furthermore, when the virtual scene generation program in memory 130 is executed by the processor, it also performs the following steps:

[0051] Based on the preset road traffic sign standards, define the virtual traffic sign model combination algorithm and the virtual traffic sign combination paradigm;

[0052] Based on the virtual traffic sign model combination algorithm and the virtual traffic sign combination paradigm, the virtual traffic sign model combination rules are obtained.

[0053] Furthermore, when the virtual scene generation program in memory 130 is executed by the processor, it also performs the following steps:

[0054] Based on the virtual traffic sign model combination algorithm and the first type of virtual traffic sign combination paradigm, at least one support model, at least one sign model, and at least one texture model are combined to generate the corresponding first type of virtual traffic sign model.

[0055] Furthermore, when the virtual scene generation program in memory 130 is executed by the processor, it also performs the following steps:

[0056] Based on the virtual traffic sign model combination algorithm and the second type of virtual traffic sign combination paradigm, at least one sign model and at least one texture model are combined to generate the corresponding second type of virtual traffic sign model.

[0057] Furthermore, when the virtual scene generation program in memory 130 is executed by the processor, it also performs the following steps:

[0058] Based on the virtual traffic sign model combination algorithm and the virtual traffic sign combination paradigm, at least one supporting object model and at least one supported object model are combined to generate a corresponding virtual traffic sign model.

[0059] Furthermore, when the virtual scene generation program in memory 130 is executed by the processor, it also performs the following steps:

[0060] Acquire road point cloud data and road line data of the target virtual scene;

[0061] The road point cloud data and the road line data are analyzed according to the preset virtual traffic sign model placement rules to determine the sign placement points corresponding to the target virtual scene. The virtual traffic sign model placement rules are obtained based on the preset road traffic sign standard definition.

[0062] Place the virtual traffic sign model at the sign placement point.

[0063] Furthermore, when the virtual scene generation program in memory 130 is executed by the processor, it also performs the following steps:

[0064] Obtain the preset requirement configuration file;

[0065] The requirement configuration file is parsed to obtain the virtual traffic sign model requirement parameters corresponding to the target virtual scene;

[0066] Based on the virtual traffic sign model requirement parameters, the steps of obtaining several pre-established virtual traffic sign component models and subsequent steps are executed.

[0067] This embodiment, through the above-described scheme, specifically acquires several pre-established virtual traffic sign component models; based on preset virtual traffic sign model combination rules, it combines these several virtual traffic sign component models to generate corresponding virtual traffic sign models. In this embodiment, the components of the virtual traffic sign model are decoupled and modeled, dividing them into several reusable virtual traffic sign component models. Based on preset virtual traffic sign model combination rules, these several virtual traffic sign component models can be combined to generate corresponding virtual traffic sign models in batches and automatically, reducing the labor costs consumed in the virtual traffic sign model generation process and improving the efficiency of virtual traffic sign model generation.

[0068] Reference Figure 2 The first embodiment of the virtual scene generation method of this application provides a flowchart, the virtual scene generation method including:

[0069] Step S10: Obtain several pre-established virtual traffic sign component models.

[0070] This embodiment relates to the generation process of a target virtual scene, which is a road virtual scene used for training and validating autonomous driving neural networks. The generation process requires a certain number of virtual traffic indicator models, which can be traffic signs, traffic lights, or other indicators used to convey specific traffic information. Furthermore, the application scenario of this solution can be extended to the generation process of road-related virtual models such as streetlights and road monitoring systems.

[0071] like Figure 3 As shown, Figure 3 This is a schematic diagram of a virtual scene, including a virtual traffic sign model, involved in the virtual scene generation method of this application. The virtual traffic sign model is one type of virtual traffic indicator model, as can be seen... Figure 3 The exhibition showcases a two-way road, including gantry-style virtual traffic sign models spanning both sides of the road, single-column virtual traffic sign models located on both sides of the road, and a single-column virtual traffic sign model located in the middle of the road. Figure 3 Each virtual traffic sign model on display includes at least three elements: a support, a signpost, and an image. It is worth noting that... Figure 3 In another scenario not shown, the virtual traffic sign model may consist only of signs and stickers, without any supporting structures. This type of virtual traffic sign model is attached to existing virtual objects on and around the road, and the attached virtual objects are called attachments.

[0072] Based on the composition rules of the virtual traffic sign model described above, sign models and texture models are pre-built, or support models are built on top of these. Support models, sign models, and texture models are all component models of the virtual traffic sign model, that is, virtual traffic sign component models.

[0073] For example, in the modeling process of support structures, single-column models are modeled for their columns, single-cantilever models are modeled for their columns and cantilever arms, and gantry models are modeled for their gantry frames. In the modeling process of signboards, based on Part 2 of the "National Standard for Road Traffic Signs and Markings" (GB5768.2—2022)—Road Traffic Signs, basic 3D signboard models of six shapes are established: rectangles (direction signs, tourist signs, road direction signs), rectangles (billboards), equilateral triangles, inverted triangles, regular octagons, and circles. In the modeling process of texture models, they are made according to the specifications of the basic 3D signboard models, ensuring at least that the size proportions are close. Texture models are essentially planar or two-dimensional models.

[0074] Similarly, for other virtual traffic sign models besides the virtual traffic sign model, virtual traffic sign component models can be pre-built to decouple the virtual traffic sign models.

[0075] At the beginning of the process of generating virtual traffic sign models, several virtual traffic sign component models are obtained. Taking the generation process of virtual traffic sign models as an example, it can be divided into the following two acquisition modes:

[0076] Acquisition Mode 1: Acquire at least one support model, at least one sign model, and at least one texture model;

[0077] Acquisition Mode 2: Acquire at least one sign model and at least one texture model.

[0078] Step S20: Based on the preset virtual traffic sign model combination rules, combine the several virtual traffic sign component models to generate the corresponding virtual traffic sign model.

[0079] Specifically, the components of the virtual traffic sign model in this embodiment are decoupled. Several acquired virtual traffic sign component models need to be combined to generate the corresponding virtual traffic sign model. Taking a virtual traffic sign model with supports as an example, at least one support model, at least one sign model, and at least one texture model are combined to generate at least one virtual traffic sign model.

[0080] The process of generating virtual traffic sign models follows preset virtual traffic sign model combination rules, which are based on relevant standards. Taking virtual traffic sign model combination rules as an example, these rules are based on road traffic sign standards, which can be set according to the "National Standard for Road Traffic Signs and Markings." Road traffic sign standards are parameterized information that can be manually set; they can be adjusted accordingly when the "National Standard for Road Traffic Signs and Markings" is updated.

[0081] Furthermore, based on the virtual traffic sign model combination rules, several virtual traffic sign component models are combined to generate the corresponding virtual traffic sign model. Taking the generation process of the virtual traffic sign model as an example, there are two specific generation modes:

[0082] Generation Mode 1: A first-type virtual traffic sign model is generated by combining at least one support model, at least one sign model, and at least one texture model. Each first-type virtual traffic sign model includes a support, a sign, and a texture. Figure 3 A type of element.

[0083] Generation Mode 2: Based on at least one sign model and at least one texture model, a corresponding second-type virtual traffic sign model is generated. Each second-type virtual traffic sign model includes two elements: a sign and a texture.

[0084] In the process of combining and generating virtual traffic sign models, the various virtual traffic sign component models are reusable, so virtual traffic sign models can be generated in batches and automatically.

[0085] This embodiment, through the above-described scheme, specifically acquires several pre-established virtual traffic sign component models; based on preset virtual traffic sign model combination rules, it combines these several virtual traffic sign component models to generate corresponding virtual traffic sign models. In this embodiment, the components of the virtual traffic sign model are decoupled and modeled, dividing them into several reusable virtual traffic sign component models. Based on preset virtual traffic sign model combination rules, these several virtual traffic sign component models can be combined to generate corresponding virtual traffic sign models in batches and automatically, reducing the labor costs consumed in the virtual traffic sign model generation process and improving the efficiency of virtual traffic sign model generation.

[0086] Furthermore, referring to Figure 4 The second embodiment of the virtual scene generation method of this application provides a flowchart, based on the above. Figure 2In the embodiment shown, before step S10, which involves obtaining several pre-established virtual traffic indicator component models, the method further includes:

[0087] Step S001: Define the virtual traffic sign model combination algorithm and the virtual traffic sign combination paradigm according to the preset road traffic sign standards.

[0088] Specifically, the virtual traffic sign model is not a random combination of several virtual traffic sign component models, but needs to meet certain virtual traffic sign model combination rules to make the generated virtual traffic sign model closer to reality and improve the effectiveness of the virtual traffic sign model for training and verification of autonomous driving neural networks.

[0089] Taking the generation process of virtual traffic sign models as an example, a preset road traffic sign standard based on the "National Standard for Road Traffic Signs and Markings" is obtained. The road traffic sign standard is a kind of parametric information. Furthermore, the virtual traffic sign model combination algorithm and virtual traffic sign combination paradigm are defined according to the road traffic sign standard. The virtual traffic sign combination paradigm can be a first type of virtual traffic sign combination paradigm or a second type of virtual traffic sign combination paradigm.

[0090] More specifically, according to the National Standard for Road Traffic Signs and Markings, road traffic signs can be roughly divided into six styles: single-column, double-column, single-cantilever, multi-cantilever, gantry, and attached.

[0091] The first type of virtual traffic sign combination paradigm includes: the first type of virtual traffic sign model = support model + sign model + texture model, corresponding to single-column, double-column, single-cantilever, multi-cantilever, and gantry virtual traffic sign models.

[0092] The second type of virtual traffic sign combination paradigm includes: second type virtual traffic sign model = sign model + sticker model, corresponding to the attached virtual traffic sign model.

[0093] The virtual traffic sign model combination algorithm includes: limiting the height of the sign model from the ground, the vertical angle of the sign model from the ground, the size of the sign model, the number of sign models, and the spacing between sign models; generating a first type of virtual traffic sign model based on the combination of three elements: support model, sign model, and texture model; and generating a second type of virtual traffic sign model based on the combination of two elements: sign model and texture model.

[0094] Taking a single-column virtual traffic sign model as an example, the role of the virtual traffic sign model combination algorithm is explained: The "National Standard for Road Traffic Signs and Markings" stipulates that the sign of a single-column traffic sign must be 2.5m to 3m above the ground, and no more than 3 signs can be embedded on one column. Therefore, the number of signs is first randomly generated within the range of 1 to 3, for example, 2; then the combination position of sign A is determined within the range of 2.5m to 3m, for example, 2.7m; then the distance between sign B and sign A is determined within the range of 0.1m to 0.5m, for example, 0.2m, and sign A is placed; finally, according to the sign and texture resource index table, two textures are randomly selected for sign A and sign B, and finally a single-column virtual traffic sign model containing two signs is generated.

[0095] Similarly, based on relevant standards for other traffic signs, corresponding virtual traffic sign model combination algorithms and virtual traffic sign combination paradigms can be further defined, and then virtual traffic sign model combination rules can be obtained based on these algorithms and paradigms. For example, when the traffic sign is a traffic light, the "Specifications for the Installation and Setting of Road Traffic Signals" (GB14886) can be used as the relevant standard.

[0096] Step S002: Based on the virtual traffic sign model combination algorithm and the virtual traffic sign combination paradigm, the virtual traffic sign model combination rules are obtained.

[0097] Specifically, after defining the virtual traffic sign model combination algorithm and the virtual traffic sign combination paradigm, the virtual traffic sign model combination algorithm and the virtual traffic sign combination paradigm are further combined to obtain the virtual traffic sign model combination rule.

[0098] Taking the generation process of virtual traffic sign models as an example, it may involve only the first type of virtual traffic sign model, or only the second type of virtual traffic sign model, or both types of virtual traffic sign models. Therefore, depending on the actual needs, corresponding virtual traffic sign model combination rules can be set, divided into the following three modes:

[0099] Mode 1: Based on the virtual traffic sign model combination algorithm and the first type of virtual traffic sign combination paradigm, the virtual traffic sign model combination rules are obtained;

[0100] Mode 2: Based on the virtual traffic sign model combination algorithm and the second type of virtual traffic sign combination paradigm, the virtual traffic sign model combination rules are obtained;

[0101] Mode 3: Based on the virtual traffic sign model combination algorithm, the first type of virtual traffic sign combination paradigm, and the second type of virtual traffic sign combination paradigm, the virtual traffic sign model combination rules are obtained.

[0102] This embodiment, through the above-described scheme, specifically defines a virtual traffic sign model combination algorithm and a virtual traffic sign combination paradigm based on preset road traffic sign standards; and obtains the virtual traffic sign model combination rules based on the virtual traffic sign model combination algorithm and the virtual traffic sign combination paradigm. This embodiment, based on relevant road traffic sign standards and component decoupling design methods, defines a virtual traffic sign model combination algorithm and a virtual traffic sign combination paradigm, and further obtains corresponding virtual traffic sign model combination rules to control the automated generation of virtual traffic sign models, thereby reducing costs and increasing efficiency in the virtual traffic sign model generation process. Furthermore, introducing road traffic sign standards can make the generated virtual traffic sign models more in line with actual needs, improving the effectiveness of virtual traffic sign models for training and validating autonomous driving neural networks.

[0103] Furthermore, referring to Figure 5 The third embodiment of the virtual scene generation method of this application provides a flowchart, based on the above. Figure 4 In the illustrated embodiment, step S20 involves combining the various virtual traffic sign component models based on preset virtual traffic sign model combination rules to generate a corresponding virtual traffic sign model. Further refinement of this model includes:

[0104] Step S201: Based on the virtual traffic sign model combination algorithm and the first type of virtual traffic sign combination paradigm, combine at least one support model, at least one sign model, and at least one texture model to generate the corresponding first type of virtual traffic sign model.

[0105] Specifically, this embodiment addresses the generation process of single-column, double-column, single-cantilever, multi-cantilever, and gantry-type virtual traffic sign models, which is the first type of virtual traffic sign model. The first type of virtual traffic sign combination paradigm includes: First type virtual traffic sign model = support model + sign model + texture model.

[0106] First, select the corresponding support model. Based on the support model, determine the number of sign models using a virtual traffic sign model combination algorithm. Then, select the corresponding sign models based on the number of sign models and combine the selected sign models with the support models. Next, query the sign and texture resource index table based on the style of the selected sign models, randomly apply textures to the selected sign models, and finally generate the first type of virtual traffic sign model.

[0107] This embodiment, through the above-described scheme, specifically by combining at least one support model, at least one sign model, and at least one texture model based on a virtual traffic sign model combination algorithm and a first-type virtual traffic sign combination paradigm, generates corresponding first-type virtual traffic sign models. This embodiment focuses on the generation process of single-column, double-column, single-cantilever, multi-cantilever, and gantry-type virtual traffic sign models. Based on the first-type virtual traffic sign combination paradigm and the virtual traffic sign model combination algorithm, the combination of support model, sign model, and texture model can automatically and in batches generate first-type virtual traffic sign models, thus reducing costs and increasing efficiency in the generation process.

[0108] Furthermore, referring to Figure 6 The fourth embodiment of the virtual scene generation method of this application provides a flowchart, based on the above. Figure 4 In the illustrated embodiment, step S20 involves combining the various virtual traffic sign component models based on preset virtual traffic sign model combination rules to generate a corresponding virtual traffic sign model. Further refinement of this model includes:

[0109] Step S202: Based on the virtual traffic sign model combination algorithm and the virtual traffic sign combination paradigm, combine at least one supporting object model and at least one supported object model to generate a corresponding virtual traffic sign model.

[0110] Specifically, this embodiment addresses the generation process of attached virtual traffic sign models, also known as the second type of virtual traffic sign models. The second type of virtual traffic sign combination paradigm includes: second type virtual traffic sign model = sign model + texture model.

[0111] First, determine the attachments corresponding to the second type of virtual traffic sign model. Based on the specifications and location of the attachments, determine the number of sign models using a virtual traffic sign model combination algorithm. Then, select the corresponding sign models based on the number of sign models and combine them with the attachments. Next, query the sign and texture resource index table based on the style of the selected sign models, randomly apply textures to the selected sign models, and finally generate the second type of virtual traffic sign model. It is worth noting that the attached virtual traffic sign model lacks a supporting model; after generating the attached virtual traffic sign model, it needs to be placed on the attached object model in the virtual scene.

[0112] This embodiment, through the above-described scheme, specifically by combining at least one sign model and at least one texture model based on a virtual traffic sign model combination algorithm and a second-type virtual traffic sign combination paradigm, generates a corresponding second-type virtual traffic sign model. This embodiment focuses on the generation process of attached virtual traffic sign models. Based on the second-type virtual traffic sign combination paradigm and the virtual traffic sign model combination algorithm, the combination of sign models and texture models can automatically and in batches generate second-type virtual traffic sign models, thus reducing costs and increasing efficiency in the generation process.

[0113] Furthermore, referring to Figure 7 The fifth embodiment of the virtual scene generation method of this application provides a flowchart, based on the above. Figure 4 In the illustrated embodiment, step S20 involves combining the various virtual traffic sign component models based on preset virtual traffic sign model combination rules to generate a corresponding virtual traffic sign model. Further refinement of this model includes:

[0114] Step S203: Based on the virtual traffic sign model combination algorithm and the virtual traffic sign combination paradigm, combine at least one supporting object model and at least one supported object model to generate a corresponding virtual traffic sign model.

[0115] This embodiment focuses on a virtual traffic indicator model composed of a support model and a supported model. Taking a traffic light as an example, the support model can correspond to a pillar, and the supported model can correspond to the light-emitting component of the traffic light. At least one support model and at least one supported model are combined to generate the corresponding virtual traffic indicator model.

[0116] This embodiment, through the above-described scheme, specifically by combining at least one supporting object model and at least one supported object model based on the virtual traffic sign model combination algorithm and the virtual traffic sign combination paradigm, generates a corresponding virtual traffic sign model. This embodiment focuses on the generation process of the virtual traffic sign model composed of supporting object models and supported object models. By combining supporting object models and supported object models based on the virtual traffic sign model combination algorithm and the virtual traffic sign combination paradigm, virtual traffic sign models can be automatically and in batches generated, thus reducing costs and increasing efficiency in the virtual traffic sign model generation process.

[0117] Furthermore, referring to Figure 8 The sixth embodiment of the virtual scene generation method of this application provides a flowchart, based on the above. Figure 2In the illustrated embodiment, after step S20, which combines the various virtual traffic sign component models based on preset virtual traffic sign model combination rules to generate a corresponding virtual traffic sign model, the method further includes:

[0118] Step S004: Obtain road point cloud data and road line data of the target virtual scene.

[0119] The target virtual scene is a type of virtual road scene, including corresponding road information. Before placing the virtual traffic sign models, it is necessary to determine the placement points of the virtual traffic sign models based on the road information.

[0120] Specifically, road information of the target virtual scene is acquired, and data extraction is performed on the road information to obtain corresponding road point cloud data and road line data. The attributes of the road point cloud data include 3D location, lane lines, road orientation, and road curvature; the attributes of the road line data include the number of lanes and road width.

[0121] Step S005: Analyze the road point cloud data and the road line data according to the preset virtual traffic sign model placement rules to determine the sign placement points corresponding to the target virtual scene. The virtual traffic sign model placement rules are based on the preset road traffic sign standard definition.

[0122] In order to place the generated virtual traffic sign models reasonably in the target virtual scene and improve the effectiveness of the target virtual scene for training and verification of autonomous driving neural networks, this embodiment defines the placement rules of virtual traffic sign models based on preset road traffic sign standards.

[0123] Specifically, taking virtual traffic sign models as an example, road traffic sign standards are pre-set according to the "National Standard for Road Traffic Signs and Markings". Road traffic sign standards are a kind of parametric information, and the placement rules of virtual traffic sign models defined therefrom also reflect the restrictions on the placement of virtual traffic sign models by the "National Standard for Road Traffic Signs and Markings".

[0124] Here are some examples illustrating the rules for placing virtual traffic sign models: ① The supports for single-column, double-column, single-cantilever, multi-cantilever, and gantry-type virtual traffic sign models cannot overlap with the roads in the target virtual scene; that is, they cannot be placed on the roads. ② Two adjacent sign models cannot be placed in positions where physical collisions may occur. ③ Single-column and double-cantilever virtual traffic sign models can be placed in the center of a two-way lane, but single-cantilever virtual traffic sign models cannot. Therefore, single-cantilever virtual traffic sign models can only be placed on the side of the oncoming or opposite lane. Single-column virtual traffic sign models are not restricted, as long as they do not physically collide with already placed virtual traffic sign models. ④ For gantry-type virtual traffic sign models, the center of the gantry must be in the center of the lane, and the width of the gantry must span both sides of the road.

[0125] Similarly, the rules for placing virtual traffic sign models can include more information about the "National Standard for Road Traffic Signs and Markings," or include the "National Standard for Road Traffic Signs and Markings."

[0126] Other basic rules besides those.

[0127] Furthermore, based on the placement rules of virtual traffic sign models, a preset sampling interval is used to sample the road point cloud data corresponding to the target virtual scene to determine the location points where virtual traffic sign models can be placed. Then, the road line data is analyzed to determine the style and orientation of the virtual traffic sign models corresponding to the location points where virtual traffic sign models can be placed. For example, if the location point where a virtual traffic sign model can be placed is on the side of the forward-facing lane, a single cantilever virtual traffic sign model can be placed with the orientation facing the forward lane.

[0128] After determining the locations where virtual traffic sign models can be placed, as well as the corresponding styles and orientations of the virtual traffic sign models, the corresponding sign placement points are obtained. Therefore, the sign placement points actually include location information, style information, and orientation information.

[0129] Similarly, the placement points of other virtual traffic indicator models besides the virtual traffic sign model can be determined using the above method.

[0130] Step S006: Place the virtual traffic sign model at the sign placement point.

[0131] Specifically, after determining the placement point of the traffic sign, the corresponding virtual traffic sign model is placed at the placement point.

[0132] like Figure 3 As shown, Figure 3 This is a schematic diagram of a virtual scene, including a virtual traffic sign model, involved in the virtual scene generation method of this application. Figure 3The exhibition showcases a two-way road, including gantry-style virtual traffic sign models spanning both sides of the road, single-column virtual traffic sign models located on both sides of the road, and a single-column virtual traffic sign model located in the middle of the road. Figure 3 The virtual traffic sign models on display are all placed at their respective sign placement points.

[0133] This embodiment, through the above-described scheme, specifically acquires road point cloud data and road line data of the target virtual scene; analyzes the road point cloud data and road line data according to preset virtual traffic sign model placement rules to determine the placement points of the signs corresponding to the target virtual scene.

[0134] The placement rules for the virtual traffic sign models are based on a preset standard definition of road traffic signs. The virtual traffic sign models are then placed at designated placement points. This embodiment introduces placement rules for virtual traffic sign models corresponding to the standard definition of road traffic signs, making the positions of the virtual traffic sign models in the virtual scene more realistic. Furthermore, by acquiring and analyzing road point cloud data and road line data of the target virtual scene, reasonable and accurate placement points for the signs can be obtained. The virtual traffic sign models are then placed at these placement points to ensure that the position, orientation, and other attributes of the virtual traffic sign models in the target virtual scene are correct, improving the robustness of the virtual traffic sign model placement process and enhancing the effectiveness of the target virtual scene for training and validating autonomous driving neural networks.

[0135] Furthermore, referring to Figure 9 The seventh embodiment of the virtual scene generation method of this application provides a flowchart, based on the above. Figure 2 In the embodiment shown, before step S10, which involves obtaining several pre-established virtual traffic indicator component models, the method further includes:

[0136] Step S007: Obtain the preset requirement configuration file.

[0137] In actual production, different virtual scenarios have different requirements for virtual traffic sign models, necessitating adjustments to the generation and placement process of these models based on specific virtual scenarios. Therefore, this embodiment introduces a requirement configuration file to control the generation and placement of virtual traffic sign models.

[0138] Specifically, a pre-defined requirements configuration file is obtained, which is edited by technical personnel. Taking the requirements configuration file corresponding to the virtual traffic sign model as an example, the content of the requirements configuration file includes: ① support model type; ② sign model type; ③ number of virtual traffic sign models.

[0139] Step S008: Parse the requirement configuration file to obtain the virtual traffic sign model requirement parameters corresponding to the target virtual scene.

[0140] Specifically, after obtaining the requirements configuration file, the requirements configuration file is parsed to obtain the corresponding virtual traffic sign model requirements parameters for the virtual scene. Taking the virtual traffic sign model as an example, based on the content of the requirements configuration file, the corresponding virtual traffic sign model requirements parameters include: ① support model type parameters; ② sign model type parameters; ③ number of virtual traffic sign models.

[0141] Step S009: Based on the virtual traffic sign model requirement parameters, perform the steps of obtaining several pre-established virtual traffic sign component models and subsequent steps.

[0142] Specifically, taking a single-column virtual traffic sign model as an example, such as Figure 10 As shown, after parsing the requirement configuration file, the corresponding virtual traffic sign model requirement parameters are obtained. Further, based on the support model type parameters, sign model type parameters, and virtual traffic sign model quantity parameters, the corresponding single-column support model and sign model can be obtained. Then, based on the sign model, the sign and texture index table is queried to obtain the corresponding texture model. The single-column support model, sign model, and texture model are combined to obtain the corresponding single-column virtual traffic sign model. Further, based on the road attributes of the target virtual scene, the corresponding sign placement points are determined, and the single-column virtual traffic sign model is placed at the corresponding sign placement points.

[0143] Similarly, for other types of virtual traffic sign models besides virtual traffic sign models, their generation and placement process can also be controlled through the corresponding requirement configuration file.

[0144] This embodiment, through the above-described scheme, specifically involves obtaining a preset requirement configuration file; parsing the requirement configuration file to obtain the virtual traffic sign model requirement parameters corresponding to the target virtual scene; and, based on the virtual traffic sign model requirement parameters, executing the steps of obtaining several pre-established virtual traffic sign component models and subsequent steps. This embodiment introduces a requirement configuration file to control the generation and placement process of virtual traffic sign models. The requirement configuration file is editable, allowing technicians to control the generation of different virtual traffic sign models by adjusting its content, meeting the needs of various virtual scenes. This effectively improves the reusability of virtual traffic sign model materials, enhances the controllability of the virtual traffic sign model generation and placement process, and further reduces costs and increases efficiency in the virtual scene generation process.

[0145] Furthermore, this application also proposes a virtual scene generation device, which includes:

[0146] The acquisition module is used to acquire several pre-established virtual traffic indicator component models;

[0147] The generation module is used to combine the several virtual traffic sign component models based on preset virtual traffic sign model combination rules to generate corresponding virtual traffic sign models.

[0148] The virtual scene generation device further includes a definition unit and a rule generation unit; the definition unit is used to define a virtual traffic sign model combination algorithm and a virtual traffic sign combination paradigm according to a preset road traffic sign standard; the rule generation unit is used to obtain the virtual traffic sign model combination rules based on the virtual traffic sign model combination algorithm and the virtual traffic sign combination paradigm.

[0149] The generation module further includes a first generation unit; the first generation unit is used to combine at least one support model, at least one sign model, and at least one texture model based on the virtual traffic sign model combination algorithm and the first type of virtual traffic sign combination paradigm to generate a corresponding first type of virtual traffic sign model.

[0150] The generation module further includes a second generation unit; the second generation unit is used to combine at least one sign model and at least one texture model based on the virtual traffic sign model combination algorithm and the second type of virtual traffic sign combination paradigm to generate a corresponding second type of virtual traffic sign model.

[0151] The generation module further includes a third generation unit; the third generation unit is used to combine at least one supporting object model and at least one supported object model based on the virtual traffic sign model combination algorithm and the virtual traffic sign combination paradigm to generate a corresponding virtual traffic sign model.

[0152] The virtual scene generation device further includes a first acquisition unit, a placement point determination unit, and a placement unit; the first acquisition unit is used to acquire road point cloud data and road line data of the target virtual scene; the placement point determination unit is used to analyze the road point cloud data and the road line data according to preset virtual traffic sign model placement rules to determine the placement points of the signs corresponding to the target virtual scene, wherein the virtual traffic sign model placement rules are obtained based on preset road traffic sign standard definitions; the placement unit is used to place the virtual traffic sign model at the placement point;

[0153] The virtual scene generation device further includes a second acquisition unit, a parsing unit, and an execution unit; the second acquisition unit is used to acquire a preset requirement configuration file; the parsing unit is used to parse the requirement configuration file to obtain the virtual traffic sign model requirement parameters corresponding to the target virtual scene; the execution unit is used to execute the steps of acquiring a number of pre-established virtual traffic sign component models and subsequent steps based on the virtual traffic sign model requirement parameters.

[0154] The principle and implementation process of virtual scene generation in this embodiment are explained in the above embodiments and will not be repeated here.

[0155] Furthermore, this application also proposes a terminal device, which includes a memory, a processor, and a virtual scene generation program stored in the memory and executable on the processor. When the virtual scene generation program is executed by the processor, it implements the steps of the virtual scene generation method described above.

[0156] Since this virtual scene generation program employs all the technical solutions of all the aforementioned embodiments when executed by the processor, it possesses at least all the beneficial effects brought about by all the technical solutions of all the aforementioned embodiments, which will not be elaborated upon here.

[0157] Furthermore, embodiments of this application also propose a computer-readable storage medium storing a virtual scene generation program, which, when executed by a processor, implements the steps of the virtual scene generation method described above.

[0158] Since this virtual scene generation program employs all the technical solutions of all the aforementioned embodiments when executed by the processor, it possesses at least all the beneficial effects brought about by all the technical solutions of all the aforementioned embodiments, which will not be elaborated upon here.

[0159] Compared to existing technologies, the virtual scene generation method, apparatus, terminal device, and storage medium proposed in this application acquire several pre-established virtual traffic sign component models; based on preset virtual traffic sign model combination rules, the several virtual traffic sign component models are combined to generate corresponding virtual traffic sign models. Based on this application's solution, the components of the virtual traffic sign model are decoupled and modeled, dividing them into several reusable virtual traffic sign component models. Based on preset virtual traffic sign model combination rules, these several virtual traffic sign component models can be combined to generate corresponding virtual traffic sign models in batches and automatically, reducing the labor costs consumed in the virtual traffic sign model generation process and improving the efficiency of virtual traffic sign model generation.

[0160] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0161] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0162] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, controlled terminal, or network device, etc.) to execute the methods of each embodiment of this application.

[0163] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for generating a virtual scene, characterized in that, The virtual scene generation method includes: Acquire several pre-established virtual traffic sign component models, wherein the virtual traffic sign component models include support models, sign models, and texture models; Based on preset virtual traffic sign model combination rules, the various virtual traffic sign component models are combined to generate corresponding virtual traffic sign models. The virtual traffic sign models include a first type of virtual traffic sign model and a second type of virtual traffic sign model. The first type of virtual traffic sign model includes a support model, a sign model, and a texture model. The second type of virtual traffic sign model includes a sign model and a texture model. The generation process of the first type of virtual traffic sign model is as follows: Select the corresponding support model; The number of sign models is determined based on a virtual traffic sign model combination algorithm; Select the appropriate sign model based on the number of sign models, and combine the selected sign model with the support model; Based on the style of the selected sign model, the sign and texture resource index table is queried, the texture model is selected, and finally the first type of virtual traffic sign model is generated. The generation process of the second type of virtual traffic sign model is as follows: Determine the attachments corresponding to the second type of virtual traffic sign model; The number of sign models is determined based on a virtual traffic sign model combination algorithm, taking into account the specifications and location of the attached objects. Select the appropriate sign model based on the number of sign models, and combine the selected sign model with the attached object; Based on the style query of the selected sign model, the sign and texture resource index table is consulted, the texture model is selected, and finally the second type of virtual traffic sign model is generated. After the step of combining the several virtual traffic sign component models based on preset virtual traffic sign model combination rules to generate the corresponding virtual traffic sign model, the method further includes: Acquire road point cloud data and road line data of the target virtual scene; Based on the placement rules of virtual traffic sign models, the road point cloud data corresponding to the target virtual scene is sampled according to the preset sampling interval to determine the location points where virtual traffic sign models can be placed. The placement rules of virtual traffic sign models are based on the preset standard definition of road traffic signs. Analyze the road line data to determine the style and orientation of the virtual traffic sign model corresponding to the location points where the virtual traffic sign model can be placed; Based on the location points where virtual traffic sign models can be placed, as well as the corresponding style and orientation of the virtual traffic sign models, the corresponding placement points for the directional signs are obtained; Place the virtual traffic sign model at the sign placement point.

2. The virtual scene generation method as described in claim 1, characterized in that, Before the step of obtaining several pre-established virtual traffic sign component models, the method further includes: Based on the preset road traffic sign standards, define the virtual traffic sign model combination algorithm and the virtual traffic sign combination paradigm; Based on the virtual traffic sign model combination algorithm and the virtual traffic sign combination paradigm, the virtual traffic sign model combination rules are obtained.

3. The virtual scene generation method as described in claim 2, characterized in that, The step of combining the several virtual traffic sign component models based on preset virtual traffic sign model combination rules to generate the corresponding virtual traffic sign model includes: Based on the virtual traffic sign model combination algorithm and the first type of virtual traffic sign combination paradigm, at least one support model, at least one sign model, and at least one texture model are combined to generate the corresponding first type of virtual traffic sign model.

4. The virtual scene generation method as described in claim 2, characterized in that, The step of combining the several virtual traffic sign component models based on preset virtual traffic sign model combination rules to generate the corresponding virtual traffic sign model includes: Based on the virtual traffic sign model combination algorithm and the second type of virtual traffic sign combination paradigm, at least one sign model and at least one texture model are combined to generate the corresponding second type of virtual traffic sign model.

5. The virtual scene generation method as described in claim 2, characterized in that, The step of combining the several virtual traffic sign component models based on preset virtual traffic sign model combination rules to generate the corresponding virtual traffic sign model includes: Based on the virtual traffic sign model combination algorithm and the virtual traffic sign combination paradigm, at least one supporting object model and at least one supported object model are combined to generate a corresponding virtual traffic sign model.

6. The virtual scene generation method as described in claim 1, characterized in that, Before the step of obtaining several pre-established virtual traffic sign component models, the method further includes: Obtain the preset requirement configuration file; The requirement configuration file is parsed to obtain the virtual traffic sign model requirement parameters corresponding to the target virtual scene; Based on the virtual traffic sign model requirement parameters, the steps of obtaining several pre-established virtual traffic sign component models and subsequent steps are executed.

7. A virtual scene generation device, characterized in that, The virtual scene generation device includes: The acquisition module is used to acquire several pre-established virtual traffic sign component models, wherein the virtual traffic sign component models include support models, sign models, and texture models; The generation module is used to combine the several virtual traffic sign component models based on preset virtual traffic sign model combination rules to generate corresponding virtual traffic sign models. The virtual traffic sign models include a first type of virtual traffic sign model and a second type of virtual traffic sign model. The first type of virtual traffic sign model includes a support model, a sign model, and a texture model. The second type of virtual traffic sign model includes a sign model and a texture model. The generation process of the first type of virtual traffic sign model is as follows: Select the corresponding support model; The number of sign models is determined based on a virtual traffic sign model combination algorithm; Select the appropriate sign model based on the number of sign models, and combine the selected sign model with the support model; Based on the style of the selected sign model, the sign and texture resource index table is queried, the texture model is selected, and finally the first type of virtual traffic sign model is generated. The generation process of the second type of virtual traffic sign model is as follows: Determine the attachments corresponding to the second type of virtual traffic sign model; The number of sign models is determined based on a virtual traffic sign model combination algorithm, taking into account the specifications and location of the attached objects. Select the appropriate sign model based on the number of sign models, and combine the selected sign model with the attached object; Based on the style query of the selected sign model, the sign and texture resource index table is consulted, the texture model is selected, and finally the second type of virtual traffic sign model is generated. After combining the various virtual traffic sign component models based on preset virtual traffic sign model combination rules to generate the corresponding virtual traffic sign model, the method further includes: Acquire road point cloud data and road line data of the target virtual scene; Based on the placement rules of virtual traffic sign models, the road point cloud data corresponding to the target virtual scene is sampled according to the preset sampling interval to determine the location points where virtual traffic sign models can be placed. The placement rules of virtual traffic sign models are based on the preset standard definition of road traffic signs. Analyze the road line data to determine the style and orientation of the virtual traffic sign model corresponding to the location points where the virtual traffic sign model can be placed; Based on the location points where virtual traffic sign models can be placed, as well as the corresponding style and orientation of the virtual traffic sign models, the corresponding placement points for the directional signs are obtained; Place the virtual traffic sign model at the sign placement point.

8. A terminal device, characterized in that, The terminal device includes a memory, a processor, and a virtual scene generation program stored in the memory and executable on the processor. When the virtual scene generation program is executed by the processor, it implements the steps of the virtual scene generation method as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a virtual scene generation program, which, when executed by a processor, implements the steps of the virtual scene generation method as described in any one of claims 1-6.

Images