Scene generation method, apparatus, and storage medium

Abstract

The application relates to a scene generation method and device, computer equipment and a storage medium. The method comprises the following steps: obtaining configuration information for a target scene to be generated; determining a target object corresponding to an object identifier, and determining at least one associated object associated with the target object; determining a selected target configuration component from a plurality of configuration components, and associating the target configuration component with the target object; determining a selected target script component from a plurality of script components, and associating the target script component with the target object; and comprehensively generating the target scene by using the target object, the at least one associated object, the target configuration component corresponding to the target object and the target script component. The method can improve the flexibility of constructing the target scene.

Description

Technical Field

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

[0002] In the field of autonomous driving simulation technology, constructing virtual scenarios for road testing can ensure the correctness of autonomous driving algorithms and help identify problems in the testing process. Therefore, constructing simulated objects in virtual scenarios and determining the interaction behavior between multiple simulated objects are key steps in road testing.

[0003] Currently, interactive rule definitions are commonly used to define and constrain simulated objects, enabling them to travel along predetermined routes, interact with the host vehicle, and ultimately reach the agreed-upon destination. However, when simulated objects move according to predefined interaction models, the interactive behavior during the movement cannot be modified temporarily. For simulated objects with complex structures, a single interaction model cannot meet the needs of road testing. Therefore, how to flexibly define the interactive behavior of different simulated objects and improve the flexibility of virtual scene generation is the problem that this application aims to solve. Summary of the Invention

[0004] Therefore, it is necessary to provide a scene generation method, apparatus, and storage medium that can improve the flexibility of constructing target scenes, addressing the aforementioned technical problems.

[0005] Firstly, this application provides a scene generation method. The method includes:

[0006] Obtain configuration information for the target scene to be generated; the configuration information includes an object identifier, multiple configuration components, and multiple script components;

[0007] Determine the target object corresponding to the object identifier, and determine at least one associated object associated with the target object;

[0008] The selected target configuration component is determined from the plurality of configuration components, and the target configuration component is associated with the target object; the target configuration component is used to indicate the behavioral characteristics of the target object in the target scene to be generated;

[0009] A selected target script component is determined from the plurality of script components, and the target script component is associated with the target object; the target script component is used to indicate the transformation feature of the target object into the at least one associated object at a target time;

[0010] By combining the target object, the at least one associated object, the target configuration component corresponding to the target object, and the target script component, a target scene is generated.

[0011] In one embodiment, the method further includes: acquiring preset map information and working condition information; the working condition information includes obstacles and traffic lights; constructing a test map for virtual testing based on the map information and the working condition information; generating a target scene by integrating the target object, the at least one associated object, the target configuration component corresponding to the target object, and the target script component, includes: performing initial configuration on the test map based on the target object and the target configuration component to obtain an initial scene corresponding to the test map; and performing detailed configuration on the initial scene based on the at least one associated object and the target script component corresponding to the target object to obtain the target scene.

[0012] In one embodiment, the target configuration component includes a transformation component, a behavior component, and a collision component; the initial configuration of the test map based on the target object and the target configuration component to obtain an initial scene corresponding to the test map includes: determining the initial position of the target object in the test map based on the transformation component, and placing the target object at the initial position; determining the behavior characteristics of the target object after starting from the initial position based on the behavior component; the behavior characteristics include at least one of straight-line behavior and avoidance behavior; configuring the target object through the collision component to obtain the initial scene; the collision component is used to indicate the collision characteristics of the target object in the target scene to be generated; the collision characteristics include at least one of simple bending and severe deformation.

[0013] In one embodiment, after initially configuring the test map according to the target object and the target configuration component to obtain the initial scene corresponding to the test map, the method further includes: determining the shadow object corresponding to the target object and obtaining the shadow configuration component corresponding to the shadow object; the shadow configuration component is used to provide an editing basis for the generation of the target configuration component; and expanding the shadow object to the initial scene according to the shadow object and the shadow configuration component to obtain the initial scene configured with the shadow object.

[0014] In one embodiment, configuring the initial scene in detail based on the at least one associated object and the target script component corresponding to the target object to obtain the target scene includes: determining the target timing when any two target objects collide based on the behavioral characteristics of the target objects configured in the initial scene; deleting the two target objects from the initial scene using the target script components corresponding to each of the two target objects; and expanding the initial scene with at least one associated object corresponding to each of the two target objects at the target timing to obtain the target scene.

[0015] In one embodiment, expanding at least one associated object corresponding to each of any two target objects into the initial scene at the target timing to obtain the target scene includes: for each of the multiple associated objects, determining the current position of the current associated object in the initial scene based on the behavior component of the target object corresponding to the current associated object, and placing the current associated object at the current position; obtaining the pre-configured attribute component of the current associated object; the attribute component is used to indicate the attribute characteristics of the associated object in the target scene to be generated; the attribute characteristics include at least the volume or weight of the current associated object; determining the current ambient temperature of the current associated object at the target timing, and determining the collision characteristics of the current associated object based on the current ambient temperature and the collision component of the target object; and combining each configured associated object to obtain the target scene.

[0016] In one embodiment, before determining at least one associated object related to the target object, the method further includes: obtaining a preset hierarchical structure of the target object and a plurality of associated objects to be associated; associating the target object with the plurality of associated objects to be associated according to the hierarchical structure to obtain the association relationship between the target object and the associated objects; determining at least one associated object related to the target object includes: determining at least one associated object related to the target object according to the association relationship.

[0017] In one embodiment, after generating the target scene, the method further includes: performing a virtual test on the target object based on the target scene to obtain test results; the test results include multiple frame test sub-results; determining the scene file corresponding to each frame test sub-result; the scene file includes at least a set of behavioral features, a set of attribute features, and a test algorithm version for each target object; and storing each scene file in a scene database.

[0018] In one embodiment, the method of performing virtual testing on the target object according to the target scene to obtain test results further includes: determining a shadow object in the target scene corresponding to the target object, and performing virtual testing on the shadow object to obtain a comparison result; performing virtual testing on the target object according to the target scene to obtain candidate results; and determining the difference between the comparison result and the candidate results to obtain a test result.

[0019] Secondly, this application also provides a scene generation apparatus. The apparatus includes:

[0020] The information acquisition module is used to acquire configuration information for the target scene to be generated; the configuration information includes an object identifier, multiple configuration components and multiple script components; determine the target object corresponding to the object identifier, and determine at least one associated object associated with the target object.

[0021] The component association module is used to determine a selected target configuration component from the plurality of configuration components and associate the target configuration component with the target object; the target configuration component is used to indicate the behavioral characteristics of the target object in the target scene to be generated; the module also determines a selected target script component from the plurality of script components and associates the target script component with the target object; the target script component is used to indicate the transformation characteristics of the target object to transform into the at least one associated object at a target time.

[0022] The scene determination module is used to generate a target scene by integrating the target object, the at least one associated object, the target configuration component and the target script component corresponding to the target object.

[0023] Thirdly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, which, when executed by a processor, performs the following steps:

[0024] Obtain configuration information for the target scene to be generated; the configuration information includes an object identifier, multiple configuration components, and multiple script components;

[0025] Determine the target object corresponding to the object identifier, and determine at least one associated object associated with the target object;

[0026] The selected target configuration component is determined from the plurality of configuration components, and the target configuration component is associated with the target object; the target configuration component is used to indicate the behavioral characteristics of the target object in the target scene to be generated;

[0027] A selected target script component is determined from the plurality of script components, and the target script component is associated with the target object; the target script component is used to indicate the transformation feature of the target object into the at least one associated object at a target time;

[0028] The target scene is generated by combining the target object, the at least one associated object, the target configuration component corresponding to the target object, and the target script component.

[0029] The aforementioned scene generation method, apparatus, and storage medium, by acquiring configuration information for the target scene to be generated, including an object identifier, multiple configuration components, and multiple script components, can determine the target object corresponding to the object identifier and at least one associated object associated with the target object. By determining the selected target configuration component from the multiple configuration components, the target configuration component can be associated with the target object; by determining the selected target script component from the multiple script components, the target script component can be associated with the target object. Thus, the target scene can be generated by combining the target object, at least one associated object, the target configuration component corresponding to the target object, and the target script component. Since this application directly associates the target configuration component and the target script component with the target object simultaneously, and the target configuration component can indicate the behavioral characteristics of the target object in the target scene to be generated, compared to the traditional method of using only a single interaction model, this application can use different target configuration components to modify the behavioral characteristics of the target object. Simultaneously, since the target script component can indicate the transformation characteristics of the target object into at least one associated object at a target time, it achieves flexible definition of the interactive behavior of different simulated objects, thereby improving the flexibility of generating virtual scenes. Attached Figure Description

[0030] Figure 1 This is an application environment diagram of the scene generation method in one embodiment;

[0031] Figure 2 This is a flowchart illustrating a scene generation method in one embodiment;

[0032] Figure 3 This is a schematic diagram of a scene editing interface in one embodiment;

[0033] Figure 4 A schematic diagram of the configuration of the target object in one embodiment;

[0034] Figure 5 This is a flowchart illustrating the process of determining the target scenario in one embodiment;

[0035] Figure 6 This is a flowchart illustrating the process of determining the target scenario in another embodiment;

[0036] Figure 7 This is a structural block diagram of a scene generation device in one embodiment;

[0037] Figure 8 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0038] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0039] The scene generation method provided in this application embodiment can be applied to, for example, Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. Terminal 102 and server 104 can execute the scene generation method provided in this embodiment of the application independently; terminal 102 and server 104 can also execute the scene generation method provided in this embodiment of the application collaboratively. Taking the collaborative execution of the scene generation method provided in this embodiment of the application by terminal 102 and server 104 as an example, terminal 102 can send the configuration information of the target scene to be generated to server 104; server 104 is used to determine the target object corresponding to the object identifier, and determine at least one associated object associated with the target object; determine the selected target configuration component from multiple configuration components, and associate the target configuration component with the target object; and determine the selected target script component from multiple script components, and associate the target script component with the target object; server 104 is also used to integrate the target object, at least one associated object, the target configuration component corresponding to the target object, and the target script component to generate the target scene, and send the target scene to terminal 102 so that terminal 102 can display the target scene. The terminal 102 may be, but is not limited to, various personal computers, laptops, smartphones, smart in-vehicle devices, tablets, IoT devices, and portable wearable devices. In one embodiment, the terminal 102 may be an autonomous driving computing platform that is not part of an autonomous driving vehicle. The server 104 may be implemented using a standalone server or a server cluster consisting of multiple servers.

[0040] In one embodiment, such as Figure 2 As shown, a scene generation method is provided. Taking the application of this method to a computer device as an example, the computer device can provide... Figure 1 The terminal or server in the process includes the following steps:

[0041] Step 202: Obtain the configuration information for the target scene to be generated.

[0042] In the virtual road test process, the autonomous driving simulation system generates several master vehicles and simulated objects with interactive behaviors based on configuration files or actual test requirements. The master vehicle is an actual or virtual vehicle controlled by the autonomous driving algorithm, while the simulated objects, also known as intelligent agents, are objects in the virtual scene that may interact with the master vehicle, such as pedestrians, motor vehicles, and bicycles. The target scene to be generated is a virtual scene that needs to include the master vehicle and simulated objects. The virtual scene contains interactions between the master vehicle and multiple simulated objects, as well as interactions between different simulated objects.

[0043] The configuration information includes an object identifier, multiple configuration components, and multiple script components. The object identifier is used to identify the target object, which can be either the main vehicle or a simulation object. The configuration components and script components each predefine different attribute characteristics, which can be used to associate with the target object, thereby configuring the target object through the attribute characteristics of different components.

[0044] Specifically, the computer device can display an information configuration interface, responding to the user's editing operations on the configuration information to obtain the configuration information of the target scene to be generated; the computer device can also export the configuration information of the target scene to be generated from a preset database.

[0045] Step 204: Determine the target object corresponding to the object identifier, and determine at least one associated object associated with the target object.

[0046] The object identifier can be an ID number, used to uniquely identify the target object.

[0047] Specifically, the computer device acquires the pre-stored association relationships between the target object and associated objects, and then determines at least one associated object related to the target object through the target object and the association relationships. The association relationships are obtained based on the hierarchical structure of the target object, which can be a tree-like node graph, i.e., having a root node and multiple child nodes connected to the root node.

[0048] In one embodiment, before determining at least one associated object associated with the target object, the method further includes: obtaining a preset hierarchical structure of the target object and multiple associated objects to be associated; and associating the target object with the multiple associated objects to be associated according to the hierarchical structure to obtain the association relationship between the target object and the associated objects.

[0049] Specifically, based on the structural characteristics of the target object, the computer device can pre-set a hierarchical structure corresponding to the target object and multiple associated objects to be associated. For example, when the target object is a pedestrian on the roadside, the root node in the hierarchical structure is the pedestrian, and the multiple child nodes are the clothes, hats, shoes, and suitcases worn by the pedestrian. When the associated objects to be associated are only the pedestrian, clothes, and shoes, the computer device associates the associated objects to be associated with the target object to obtain the association relationship between the target object and the associated objects. As another example, if the target object has a long-tail structure, that is, when the target object is a trailer carrying a bicycle, the root node in the hierarchical structure is the trailer, and the child nodes are the bicycles. Therefore, when the associated objects to be associated are the trailer and the bicycle, the computer device associates the bicycle and the trailer with the target object to obtain the association relationship between the target object and the associated objects.

[0050] Among them, the target object of the long-tail structure is a complex object. Complex objects will exhibit a long-tail problem in the field of autonomous driving simulation, such as a trailer with a non-moving vehicle, divided into several sections, or a train that cannot be regarded as such. A single interaction model cannot meet the interaction behavior of complex objects.

[0051] In this embodiment, the relationship between the target object and related objects is constructed through different hierarchical structures, which enables flexible definition of the target object. At the same time, simple related objects can be constructed into complex objects with complex structures to meet the needs of different virtual scenes, thus ensuring the realism of the virtual scene construction.

[0052] In one embodiment, the target object can be viewed as including the associated objects corresponding to the root node and the associated objects corresponding to the child nodes. The associated objects corresponding to the root node can be referred to as the root object, and the associated objects corresponding to the child nodes can be referred to as child objects. The attribute characteristics of the child objects are influenced by the attribute characteristics of the root object, but some attribute characteristics of the child objects are still determined by their own model type. Referring to the example above, if the root object is a pedestrian and the child objects are clothes and shoes, when the person walks wearing shoes, the child objects will move along with the root object, but will have their own independent attribute characteristics, such as the shoes wearing down with movement. As another example, if the root object is a cart and the child object is a bicycle, when the cart moves forward, the child objects will move along with it, but when the bicycle moves back and forth on the cart, the root object will not move.

[0053] Step 206: Determine the selected target configuration component from multiple configuration components and associate the target configuration component with the target object.

[0054] The target configuration component is used to indicate the behavioral characteristics of the target object in the target scene to be generated; the target configuration component also includes a model component, which is used to define the model type of the target object, for example, the model type of the target object is pedestrian, bicycle, or car.

[0055] Specifically, the computer device can display an information configuration interface including multiple configuration components. When the computer device responds to a user's selection operation on these components, it determines the target configuration component selected. These configuration components are stored in a predefined component library. For model components within the target configuration component, the computer device can import them from a third-party model library. When the computer device associates the target configuration component with a target object, the target object can be configured with all the attribute characteristics of the target configuration component.

[0056] In one embodiment, before associating a target object with a target configuration component, the computer device may treat the target object as an empty object, that is, without any attribute characteristics. The interactive behavior of the target object in the target scene to be generated will also be different depending on the components mounted in the empty object.

[0057] Step 208: Determine the selected target script component from multiple script components and associate the target script component with the target object.

[0058] The target script component is used to indicate the transformation characteristics of a target object into at least one associated object at a target time. The target time usually refers to the timing of certain sudden events encountered by the target object, such as a collision between any two target objects. The transformation characteristics can characterize the different behaviors of the associated objects based on their own attributes after the target object is transformed into at least one associated object. For example, when a wall is hit, it will be transformed into multiple wall blocks, and the wall blocks of different weights will be scattered in different locations.

[0059] Specifically, the computer device can display an information configuration interface, which includes multiple script components. When the computer device responds to a user's selection operation on multiple script components, it determines the target script component selected by the selection operation. These script components are stored in a predefined component library. After the computer device associates a target object with a target configuration component, the target object can be considered to have basic attribute characteristics configured. After associating a target object with a target script component, the target object can be considered to have complete attribute characteristics configured. At this point, the attribute characteristics in the target script component can be represented by a separate script file. For example, when a target object is configured with a separate script file, and a target object collides with a fan, the separate script file can control the transformation of the target object and determine its transformation characteristics when it is transformed into at least one associated object.

[0060] Step 210: Combine the target object, at least one associated object, the target configuration component corresponding to the target object, and the target script component to generate the target scene.

[0061] In one embodiment, preset map information and working condition information are obtained; a test map for virtual testing is constructed based on the map information and working condition information; a target scene is generated by integrating the target object, at least one associated object, the target configuration component corresponding to the target object, and the target script component, including: initial configuration of the test map based on the target object and the target configuration component to obtain the initial scene corresponding to the test map; detailed configuration of the initial scene based on at least one associated object and the target script component corresponding to the target object to obtain the target scene.

[0062] The map information may include road markings, such as road areas, lane line types, and building areas. The map information can be obtained through user-preset road arrangement rules or edited and generated through an editor system. The operational information includes obstacles, traffic lights, and road signs.

[0063] Specifically, such as Figure 3 As shown, Figure 3 This is a schematic diagram of a scene editing interface. The computer device constructs a test map for virtual testing based on map information and operational information, responding to user editing operations within the scene editing interface. The computer device determines the location and behavior information of the target object based on the target configuration component, and places the target object at the initial position indicated by the location information on the test map. Then, based on the location and behavior information, it performs initial configuration of the test map to obtain the initial scene corresponding to the test map. At this point, the target object in the initial scene has certain behavioral characteristics, such as avoiding obstacles. After obtaining the initial scene, the computer device can perform detailed configuration of the initial scene based on the target script component and at least one associated object, that is, configure the transformation characteristics of the target object into the associated object at the target time, thus obtaining the target scene.

[0064] In one embodiment, the attribute features in the target configuration component may further include initial speed and direction of travel, thereby ensuring that the target object starts traveling in the target scene according to the initial speed and direction of travel.

[0065] The aforementioned scene generation method obtains configuration information for the target scene to be generated. This configuration information includes an object identifier, multiple configuration components, and multiple script components. This allows for the identification of the target object corresponding to the object identifier and the determination of at least one associated object. By selecting a target configuration component from the multiple configuration components, the target configuration component can be associated with the target object. Similarly, by selecting a target script component from the multiple script components, the target script component can be associated with the target object. Thus, the target scene is generated by combining the target object, at least one associated object, the target configuration component corresponding to the target object, and the target script component. Since this application directly associates both the target configuration component and the target script component with the target object, and the target configuration component indicates the behavioral characteristics of the target object in the target scene to be generated, compared to the traditional method of using only a single interaction model, this application can use different target configuration components to modify the behavioral characteristics of the target object. Furthermore, since the target script component indicates the transformation characteristics of the target object into at least one associated object at a target time, it enables flexible definition of the interactive behavior of different simulated objects, thereby improving the flexibility of virtual scene generation.

[0066] In one embodiment, the target configuration component includes a transformation component, a behavior component, and a collision component; based on the target object and the target configuration component, the test map is initially configured to obtain the initial scene corresponding to the test map, including: determining the initial position of the target object in the test map based on the transformation component and placing the target object at the initial position; determining the behavior characteristics of the target object after starting from the initial position based on the behavior component; configuring the target object through the collision component to obtain the initial scene; the collision component is used to indicate the collision characteristics of the target object in the target scene to be generated.

[0067] The transformation component represents the initial position of the target object on the test map. The behavior component can include a straight-moving component and an avoidance component. The straight-moving component represents that the target object will move forward in a straight line when there are no factors such as traffic lights. The avoidance component represents that when encountering obstacles, the target object will avoid them based on specific historical information, such as deceleration avoidance or lane change avoidance. The collision component can represent whether the target object is a rigid body and whether it is transformed into a suitable physical model upon collision. The collision feature corresponding to the collision component includes at least one of simple bending and severe deformation.

[0068] Specifically, such as Figure 4 As shown, Figure 4To configure the schematic diagram of the target object, the computer device determines the initial position of the target object on the test map based on the transformation component, places the target object at the initial position, and then determines the behavioral characteristics of the target object after starting from the initial position based on the behavior component. These behavioral characteristics include at least one of straight-line behavior and avoidance behavior; straight-line behavior corresponds to the straight-line component, and avoidance behavior corresponds to the avoidance component. For example, [the following is an example of a specific configuration]. Figure 4 When the straight-line component and the transformation component are connected to object 2, it indicates that the behavior of object 2 is to start from the initial position and move forward in a straight line. The computer device configures the target object through the collision component, that is, connects the collision component to the target object to obtain the initial scene. For example, [the following text is incomplete and likely refers to a different scene:] Figure 4 When the collision component and the model component with the bicycle are connected to object 2, it indicates that object 2 is a bicycle moving forward and will collide. At this time, object 2 is the target object.

[0069] In one embodiment, to avoid conflicts in the interaction behaviors of different target objects, different priority orders can be marked for behavior components. For example, the priority of the avoidance component is higher than that of the straight-through component.

[0070] In this embodiment, by using the transformation component, behavior component and collision component in the target configuration component to construct the behavioral characteristics of the target object in the test map, the configuration process of the target object with a complex structure can be simplified, and the interactive behavior of the target object can be made more in line with the interactive logic in the virtual scene. This avoids the problem of wasting autonomous driving simulation resources and increasing maintenance costs due to the excessive introduction of too many target objects with complex structures.

[0071] In one embodiment, after initially configuring the test map based on the target object and the target configuration component to obtain the initial scene corresponding to the test map, the method further includes: determining the shadow object corresponding to the target object and obtaining the shadow configuration component corresponding to the shadow object; the shadow configuration component is used to provide an editing basis for the generation of the target configuration component; based on the shadow object and the shadow configuration component, the shadow object is extended to the initial scene to obtain the initial scene configured with the shadow object.

[0072] In cases where the target scene to be generated involves interaction between the main vehicle and simulated objects, and the main vehicle's behavior does not conform to the interaction logic in the virtual scene, the behavior of the main vehicle needs to be modified to ensure that the modified main vehicle can interact normally with the simulated objects. If the original main vehicle still needs to be displayed in the target scene to be generated, it will not interact with other simulated objects, but it can be compared with the modified main vehicle to determine the differences between the original and modified test algorithm versions. Here, the original main vehicle is equivalent to a shadow object, and the modified main vehicle is equivalent to the target object.

[0073] Specifically, since shadow objects do not interact with other simulated objects, their shadow configuration components typically include transformation and behavior components, but not collision components, target script components, etc., associated with the target object. (See reference...) Figure 4 , Figure 4 Object 3 in the image represents a shadow mode. Furthermore, since the computer device redefines the target object when the shadow object does not conform to the interaction logic in the virtual scene, the target configuration component is obtained based on the shadow configuration component. The specific implementation process of associating the shadow configuration component with the shadow object can be found in the specific implementation process of associating the target configuration component with the target object, and will not be repeated here. Similarly, the specific implementation process of extending the shadow object to the initial scene can be found in the specific implementation process of initial configuration of the test map based on the target object and the target configuration component, and will not be repeated here.

[0074] In this embodiment, by defining a shadow object in the target scene to be generated, it is possible to compare it with the corresponding target object, thereby determining the differences between different test algorithm versions.

[0075] In one embodiment, such as Figure 5 As shown, based on at least one associated object and the target script component corresponding to the target object, the initial scene is configured in detail to obtain the target scene, including the following steps:

[0076] Step 502: Determine the target timing when any two target objects collide, based on the behavioral characteristics of the target objects configured in the initial scene.

[0077] Since different target objects have different target configuration components, when different target objects move forward along different routes in the initial scene, they may collide at a certain position. Therefore, the computer device takes the time when any two target objects collide as the target timing.

[0078] Step 504: Delete any two target objects from the initial scene using the target script components corresponding to each of the two target objects.

[0079] Specifically, when any two target objects are configured with their respective corresponding target script components, it indicates that the two target objects will produce different behaviors based on their own attribute characteristics. That is, the computer device will convert each target object into at least one corresponding associated object and delete any two target objects from the initial scene. For example, if target object 1 is a person pushing a bicycle and target object 2 is a wall, when target object 1 and target object 2 collide, the computer device will convert target object 1 into a bicycle and a pedestrian, and delete the person pushing the bicycle from the initial scene; similarly, target object 2 will be converted into multiple wall blocks and the walls will be deleted.

[0080] Step 506: At the target timing, expand at least one associated object corresponding to each of the two target objects into the initial scene to obtain the target scene.

[0081] Furthermore, such as Figure 6 As shown, at the target timing, at least one associated object corresponding to each of any two target objects is expanded into the initial scene to obtain the target scene, including the following steps:

[0082] Step 602: For each of the multiple associated objects, determine the current position of the current associated object in the initial scene based on the behavior component of the target object corresponding to the current associated object, and place the current associated object at the current position.

[0083] Specifically, since the target object has already traveled to a different position than its initial position when the collision occurs, the computer device needs to determine the collision position through the behavior component of the target object corresponding to the currently associated object before deleting the target object from the initial scene. The collision position is then regarded as the current position of the currently associated object in the initial scene, so that the computer device can place the currently associated object at the current position.

[0084] Step 604: Obtain the pre-configured attribute components of the currently associated object.

[0085] The attribute component is used to indicate the attribute characteristics of the associated object in the target scene to be generated. The computer device can find the attribute characteristics of the associated object through a pre-set index method. The attribute characteristics include at least the volume or weight of the current associated object. Referring to the example above, when the target object 1 is converted into a bicycle and a pedestrian, the attribute characteristics can be the bicycle model and the pedestrian model, or the attribute characteristics can be the weight of each wall block, etc.

[0086] Step 606: Determine the current ambient temperature of the current associated object at the target time, and determine the collision characteristics of the current associated object based on the current ambient temperature and the collision components of the target object.

[0087] Specifically, the computer device triggers the detection components configured on the target object. These components detect the ambient temperature to determine the current ambient temperature of the associated object at the target time. Since each target object can be equipped with multiple different components, the collision characteristics triggered by the collision components will differ depending on the current ambient temperature. For example, if the current ambient temperature is high, the collision characteristic of the bicycle model is simple bending; conversely, if the ambient temperature is low, the collision characteristic is severe deformation.

[0088] Step 608: Combine the associated objects that have been configured to obtain the target scenario.

[0089] In this embodiment, the target object can be controlled through the target script component, and the transformation characteristics of the target object into at least one associated object at the target time can be determined. Since the collision characteristics of different current ambient temperatures can be determined through the collision component for each associated object, the different attribute characteristics of the associated object can be flexibly determined.

[0090] In one embodiment, the flowchart for determining the target model, after generating the target scene, further includes: performing a virtual test on the target object based on the target scene to obtain test results; the test results include multiple frame test sub-results; determining the scene file corresponding to each frame test sub-result; and storing each scene file in the scene database.

[0091] The scene file includes at least the behavioral feature set, attribute feature set, test algorithm version, map information used, and whether historical data collected from roads is used for each target object. The scene file represents a description file for simulating the target scene, which records the configuration information of the target objects and the interaction information between different target objects.

[0092] Specifically, after the target scene is generated, the computer device can perform virtual testing on the target object using a simulator. The simulator is a core program with a simulation algorithm that uses a predefined description file associated with the target scene as initial conditions for the virtual test. Since each target object undergoes a state change at a preset interval, for example, 0.1 seconds, the virtual test results include multiple frames of test sub-results. The computer device can determine the scene file corresponding to each of the multiple test sub-results, that is, determine the running state of the target object at different times, the interaction state between different target objects, and other information.

[0093] In this embodiment, since the generated target scene has a certain degree of realism, each target object uses different components for state updates, ensuring that the target scene achieves a balance between diversity and meeting the expected effect, which can meet the needs of different complex types of autonomous driving simulation tests.

[0094] In one embodiment, the method further includes: performing a virtual test on the target object according to the target scene to obtain test results; determining a shadow object in the target scene corresponding to the target object and performing a virtual test on the shadow object to obtain a comparison result; performing a virtual test on the target object according to the target scene to obtain candidate results; and obtaining a test result based on the difference between the comparison result and the candidate results.

[0095] Specifically, the defined shadow object carries a shadow configuration component. When the computer device determines the shadow object corresponding to the target object, it can perform a virtual test on the shadow object to obtain a comparison result, and simultaneously perform a virtual test on the target object to obtain candidate results. The computer device determines the difference between the comparison result and the candidate results, that is, determines the difference between different versions of the test algorithm, and uses the difference as the test result. Furthermore, since the target configuration component of the target object is modified and edited based on the shadow configuration component, the correctness of the modified test algorithm can be determined based on the test results. Therefore, this embodiment ensures the accuracy of the target object's interactive behavior by determining the simulation test differences between the shadow object and the target object at different times.

[0096] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0097] Based on the same inventive concept, this application also provides a scene generation apparatus for implementing the scene generation method described above. The solution provided by this apparatus is similar to the implementation scheme described in the above method; therefore, the specific limitations in one or more scene generation apparatus embodiments provided below can be found in the limitations of the scene generation method described above, and will not be repeated here.

[0098] In one embodiment, such as Figure 7As shown, a scene generation device 700 is provided, including: an information acquisition module 702, a component association module 704, and a scene determination module 706, wherein:

[0099] The information acquisition module 702 is used to acquire configuration information for the target scene to be generated; the configuration information includes an object identifier, multiple configuration components and multiple script components; the target object corresponding to the object identifier is determined, and at least one associated object associated with the target object is determined.

[0100] The component association module 704 is used to determine the selected target configuration component from multiple configuration components and associate the target configuration component with the target object; the target configuration component is used to indicate the behavioral characteristics of the target object in the target scene to be generated; the module also determines the selected target script component from multiple script components and associates the target script component with the target object; the target script component is used to indicate the transformation characteristics of the target object when it transforms into at least one associated object at the target time.

[0101] The scene determination module 706 is used to generate a target scene by integrating the target object, at least one associated object, the target configuration component corresponding to the target object, and the target script component.

[0102] In one embodiment, the scenario determination module 706 is further configured to acquire preset map information and working condition information; construct a test map for virtual testing based on the map information and working condition information; and generate a target scenario by integrating the target object, at least one associated object, the target configuration component corresponding to the target object, and the target script component, including: performing initial configuration on the test map based on the target object and the target configuration component to obtain an initial scenario corresponding to the test map; and performing detailed configuration on the initial scenario based on at least one associated object and the target script component corresponding to the target object to obtain the target scenario.

[0103] In one embodiment, the scene determination module 706 further includes an initial configuration module 7061, which is used to determine the initial position of the target object in the test map according to the transformation component and place the target object at the initial position; determine the behavior characteristics of the target object after starting from the initial position according to the behavior component; the behavior characteristics include at least one of straight-line behavior and avoidance behavior; configure the target object through the collision component to obtain the initial scene; the collision component is used to indicate the collision characteristics of the target object in the target scene to be generated; the collision characteristics include at least one of simple bending and severe deformation.

[0104] In one embodiment, the scene determination module 706 further includes a shadow component configuration module 7062, which is used to determine the shadow object corresponding to the target object and obtain the shadow configuration component corresponding to the shadow object; the shadow configuration component is used to provide an editing basis for the generation of the target configuration component; based on the shadow object and the shadow configuration component, the shadow object is extended to the initial scene to obtain the initial scene configured with the shadow object.

[0105] In one embodiment, the scene determination module 706 further includes a detailed configuration module 7063, which is used to determine the target timing when any two target objects collide based on the behavioral characteristics of the target objects configured in the initial scene; delete any two target objects from the initial scene through the target script components corresponding to each of the two target objects; and expand at least one associated object corresponding to each of the two target objects into the initial scene at the target timing to obtain the target scene.

[0106] In one embodiment, the detailed configuration module 7063 is further configured to, for each of the multiple associated objects, determine the current position of the current associated object in the initial scene based on the behavior component of the target object corresponding to the current associated object, and place the current associated object at the current position; obtain the pre-configured attribute component of the current associated object; the attribute component is used to indicate the attribute characteristics of the associated object in the target scene to be generated; determine the current ambient temperature of the current associated object at the target time, and determine the collision characteristics of the current associated object based on the current ambient temperature and the collision component of the target object; and synthesize each configured associated object to obtain the target scene.

[0107] In one embodiment, the scene generation device 700 further includes an object association module 708, which is used to obtain a preset hierarchical structure of the target object and a plurality of associated objects to be associated; according to the hierarchical structure, associate the target object with the plurality of associated objects to be associated to obtain the association relationship between the target object and the associated objects; and determine at least one associated object associated with the target object, including: determining at least one associated object associated with the target object according to the association relationship.

[0108] In one embodiment, the scene generation device 700 further includes a scene testing module 710, which is used to perform virtual testing on the target object according to the target scene and obtain test results; the test results include multiple frame test sub-results; determine the scene file corresponding to each frame test sub-result; the scene file includes a set of behavioral features, a set of attribute features and a test algorithm version of each target object; and store each scene file in a scene database.

[0109] In one embodiment, the scene testing module 710 is further configured to determine the shadow object corresponding to the target object in the target scene, and perform virtual testing on the shadow object to obtain a comparison result; perform virtual testing on the target object according to the target scene to obtain a candidate result; determine the difference between the comparison result and the candidate result to obtain a test result.

[0110] Each module in the aforementioned scene generation device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0111] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 8 As shown, the computer device includes a processor, memory, input / output interfaces (I / O), a communication interface, a display unit, and input devices. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface, display unit, and input devices are also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a scene generation method. The display unit of the computer device is used to form a visually visible image. It can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0112] Those skilled in the art will understand that Figure 8 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0113] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0114] In one embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0115] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0116] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0117] The above embodiments are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A scene generation method characterized by, The method comprises: obtaining configuration information for a target scene to be generated; the configuration information comprises an object identifier, a plurality of configuration components and a plurality of script components; determining a target object corresponding to the object identifier, and determining at least one associated object associated with the target object; determining a target configuration component selected from the plurality of configuration components, and associating the target configuration component with the target object; the target configuration component is used to indicate a behavior feature of the target object in the target scene to be generated; determining a target script component selected from the plurality of script components, and associating the target script component with the target object; the target script component is used to indicate a conversion feature of the target object into the at least one associated object at a target time; integrating the target object, the at least one associated object, the target configuration component corresponding to the target object and the target script component to generate a target scene; before the determining the at least one associated object associated with the target object, the method further comprises: obtaining a preset hierarchical structure of the target object and a plurality of associated objects to be associated; and associating the target object with the plurality of associated objects to be associated according to the hierarchical structure to obtain an association relationship between the target object and the associated objects.

2. The method of claim 1, wherein, The method further comprises: obtaining preset map information and working condition information; the working condition information at least comprises an obstacle and a traffic light; constructing a test map for virtual testing according to the map information and the working condition information; The integrating the target object, the at least one associated object, the target configuration component corresponding to the target object and the target script component to generate a target scene comprises: performing initial configuration on the test map according to the target object and the target configuration component to obtain an initial scene corresponding to the test map; performing detailed configuration on the initial scene according to the at least one associated object and the target script component corresponding to the target object to obtain a target scene.

3. The method of claim 2, wherein, The target configuration component comprises a transformation component, a behavior component and a collision component; the performing initial configuration on the test map according to the target object and the target configuration component to obtain an initial scene corresponding to the test map comprises: determining an initial position of the target object in the test map according to the transformation component, and placing the target object at the initial position; determining a behavior feature of the target object after starting from the initial position according to the behavior component; the behavior feature at least comprises one of a straight-line behavior and an avoidance behavior; performing configuration on the target object through the collision component to obtain an initial scene; the collision component is used to indicate a collision feature of the target object in the target scene to be generated; the collision feature at least comprises one of a simple bending and a serious deformation.

4. The method of claim 3, wherein, After the performing initial configuration on the test map according to the target object and the target configuration component to obtain an initial scene corresponding to the test map, the method further comprises: determine a shadow object corresponding to the target object, and obtain a shadow configuration component corresponding to the shadow object; the shadow configuration component is used to provide an editing basis for generation of the target configuration component; the shadow object is expanded into the initial scene according to the shadow object and the shadow configuration component, to obtain the initial scene configured with the shadow object.

5. The method of claim 2, wherein, the initial scene is configured in detail according to a target script component corresponding to the at least one associated object and the target object, to obtain a target scene, including: determine a target timing when any two target objects collide according to a behavior feature of the target object configured in the initial scene; delete the any two target objects from the initial scene through the target script component corresponding to each of the any two target objects; expand at least one associated object corresponding to each of the any two target objects into the initial scene at the target timing, to obtain a target scene.

6. The method of claim 5, wherein, the at least one associated object corresponding to each of the any two target objects is expanded into the initial scene at the target timing, to obtain a target scene, including: for each associated object in the plurality of associated objects, determine a current position of the current associated object in the initial scene according to a behavior component of the target object corresponding to the current associated object, and place the current associated object at the current position; obtain a property component of the current associated object configured in advance; the property component is used to indicate a property feature of an associated object in a target scene to be generated; the property feature at least includes a volume or a weight of the current associated object; determine a current environmental temperature of the current associated object at the target timing, and determine a collision feature of the current associated object according to the current environmental temperature and a collision component of the target object; obtain a target scene by integrating each of the associated objects configured.

7. The method of claim 1, wherein the at least one associated object associated with the target object is determined, including: determine the at least one associated object associated with the target object according to the association relationship.

8. The method of claim 1, wherein, after the target scene is generated, the method further includes: virtually test the target object according to the target scene, to obtain a test result; the test result includes a plurality of test sub-results; determine a scene file corresponding to each test sub-result; the scene file at least includes a behavior feature set, a property feature set and a test algorithm version of each target object; store each scene file into a scene database.

9. The method of claim 8, wherein, the target object is virtually tested according to the target scene, to obtain a test result, and the method further includes: determine a shadow object corresponding to the target object in the target scene, and virtually test the shadow object, to obtain a comparison result; virtually test the target object according to the target scene, to obtain a candidate result; obtain a test result according to a difference between the comparison result and the candidate result.

10. A scene generation apparatus characterized by comprising: the device includes: An information obtaining module is configured to obtain configuration information for a target scene to be generated; the configuration information comprises an object identifier, a plurality of configuration components, and a plurality of script components; determine a target object corresponding to the object identifier, and determine at least one associated object associated with the target object; A component associating module is configured to determine a target configuration component selected from the plurality of configuration components, and associate the target configuration component with the target object; the target configuration component is used to indicate a behavior feature of the target object in the target scene to be generated; determine a target script component selected from the plurality of script components, and associate the target script component with the target object; the target script component is used to indicate a conversion feature of the target object into the at least one associated object at a target time; A scene determining module is configured to generate a target scene by comprehensively considering the target object, the at least one associated object, the target configuration component corresponding to the target object, and the target script component. An object associating module is configured to obtain a preset hierarchical structure of the target object and a plurality of associated objects to be associated; and associate the target object with the plurality of associated objects to be associated according to the hierarchical structure, to obtain an association relationship between the target object and the associated objects.

11. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program, when executed by a processor, implements the steps of the method of any one of claims 1 to 9.

Images