Method and device for constructing scene restoration data, vehicle and storage medium

Abstract

The application relates to a method and device for constructing scene restoration data, a vehicle and a storage medium, wherein the method for constructing scene restoration data comprises the following steps: obtaining test perception data, test vehicle-road cooperation data and high-precision map data of a target scene; respectively segmenting the test perception data, the test vehicle-road cooperation data and the high-precision map data according to the same rule to determine a plurality of test perception data segments, a plurality of test vehicle-road cooperation data segments and a plurality of high-precision map data segments; performing fusion processing on the test perception data segments, the test vehicle-road cooperation data segments and the high-precision map data segments corresponding to each other to determine scene restoration data segments; and determining the scene restoration data of the target scene according to the plurality of scene restoration data segments. The application can improve the data stability, data accuracy and data predictability in the restored scene, and improve the use experience of scene restoration.

Description

Technical Field

[0001] This application relates to the field of vehicle technology, specifically to a method, apparatus, vehicle, and storage medium for constructing scene reconstruction data. Background Technology

[0002] High-precision maps, also known as autonomous driving maps or high-resolution maps, are a new map data paradigm for autonomous vehicles. They generally consist of three types of vector information: lane models containing semantic information, road components, and road attributes, as well as feature layers for multi-sensor localization.

[0003] In autonomous vehicles, high-definition maps are typically used to reconstruct driving scenarios, enabling driving prediction and achieving autonomous driving. However, the coverage area of ​​current high-definition maps is limited, and they cannot reconstruct relevant scenarios in uncovered areas, resulting in a poor user experience. Summary of the Invention

[0004] One objective of this application is to provide a method for constructing scene reconstruction data, which can obtain scene reconstruction data with better reliability, higher richness and wider coverage; a second objective of this application is to provide a driving prediction method; a third objective of this application is to provide an apparatus for constructing scene reconstruction data; a fourth objective of this application is to provide a driving prediction apparatus; a fifth objective of this application is to provide a vehicle; and a sixth objective of this application is to provide a storage medium.

[0005] To achieve the above objectives, firstly, this application provides a method for constructing scene reconstruction data, the method comprising:

[0006] Acquire test perception data, test vehicle-road cooperative data, and high-precision map data for the target scenario;

[0007] The test perception data, the test vehicle-road cooperative data, and the high-precision map data are segmented according to the same rules to determine multiple test perception data segments corresponding to the test perception data, multiple test vehicle-road cooperative data segments corresponding to the test vehicle-road cooperative data, and multiple high-precision map data segments corresponding to the high-precision map data; wherein, the multiple test perception data segments, the multiple test vehicle-road cooperative data segments, and the multiple high-precision map data segments correspond one-to-one.

[0008] The corresponding test perception data segment, the test vehicle-road cooperative data segment, and the high-precision map data segment are fused to determine the scene reconstruction data segment;

[0009] Based on multiple scene restoration data segments, the scene restoration data of the target scene is determined.

[0010] Further, the fusion processing of the corresponding test perception data segment, the test vehicle-road cooperative data segment, and the high-precision map data segment to determine the scene reconstruction data segment includes:

[0011] The data in the test perception data segment, the test vehicle-road cooperative data segment, and the high-precision map data segment are respectively processed into images in the form of data points to determine the perception image corresponding to the test perception data segment, the vehicle-road cooperative image corresponding to the test vehicle-road cooperative data segment, and the high-precision map image corresponding to the high-precision map data segment.

[0012] Perform difference analysis on the perceived image, the vehicle-road cooperative image, and the high-precision map image;

[0013] Based on the difference analysis results, the scene restoration data segment is selected from the test perception data segment, the test vehicle-road cooperative data segment, and the high-precision map data segment.

[0014] Furthermore, the selection of the scene reconstruction data segment from the test perception data segment, the test vehicle-road cooperative data segment, and the high-precision map data segment based on the difference analysis results includes at least one of the following:

[0015] For a first region where there are identical data points among the perceived image, the vehicle-road cooperative image, and the high-precision map image, select the scene restoration data of the first region corresponding to the first region from any one of the test perception data segment, the test vehicle-road cooperative data segment, and the high-precision map data segment;

[0016] For the second region where data points exist in one of the perception images, the vehicle-road cooperative images, and the high-precision map images, the scene restoration data of the second region corresponding to the second region is selected from the data segment corresponding to the image where data points exist in the second region.

[0017] For the third region in the perception image, the vehicle-road cooperative image and the high-precision map image, where two images have the same data points, the scene restoration data of the third region corresponding to the third region is selected from the data segment corresponding to either of the two images;

[0018] For a fourth region in which at least two of the perceived image, the vehicle-road cooperative image, and the high-precision map image have data points and the data points differ, the scene restoration data of the fourth region corresponding to the fourth region is selected from the data segments with higher priority among the data segments corresponding to the at least two images.

[0019] Furthermore, the priority of the test perception data segment is higher than the priority of the test vehicle-road cooperative data segment, and the priority of the test vehicle-road cooperative data segment is higher than the priority of the high-precision map data segment.

[0020] Further, the step of segmenting the test perception data, the test vehicle-road cooperative data, and the high-precision map data according to the same rules to determine multiple test perception data segments corresponding to the test perception data, multiple test vehicle-road cooperative data segments corresponding to the test vehicle-road cooperative data, and multiple high-precision map data segments corresponding to the high-precision map data includes:

[0021] The target scene is divided into multiple sub-regions based on a set size.

[0022] The test perception data, the test vehicle-road cooperative data, and the high-precision map data are divided into multiple sub-regions to determine multiple test perception data segments, multiple test vehicle-road cooperative data segments, and multiple high-precision map data segments.

[0023] Furthermore, the sensory data used for testing is obtained in the following manner:

[0024] Acquire image data of the target scene;

[0025] The image data is converted to determine the sensory data for testing.

[0026] To achieve the above objectives, secondly, this application also provides a driving prediction method, the driving prediction method comprising:

[0027] Obtain the original scene reconstruction data of the target scene; wherein the original scene reconstruction data is determined by the method for constructing scene reconstruction data as described in the first aspect;

[0028] Based on the original scene reconstruction data, scene prediction is performed for driving in the target scene.

[0029] Furthermore, the step of performing scene prediction for driving in the target scene based on the original scene reconstruction data includes:

[0030] Acquire driving perception data and / or driving vehicle-to-infrastructure (V2I) data;

[0031] Based on the driving perception data and / or the driving vehicle-road cooperative data, as well as the original scene reconstruction data, the target scene reconstruction data is constructed.

[0032] Based on the target scene reconstruction data, scene prediction is performed for driving in the target scene.

[0033] To achieve the above objectives, in a third aspect, this application also provides an apparatus for constructing scene reconstruction data, the apparatus comprising:

[0034] The first acquisition module is used to acquire test perception data, test vehicle-road cooperative data and high-precision map data of the target scene;

[0035] The determination module is used to segment the test perception data, the test vehicle-road cooperative data, and the high-precision map data according to the same rules, so as to determine multiple test perception data segments corresponding to the test perception data, multiple test vehicle-road cooperative data segments corresponding to the test vehicle-road cooperative data, and multiple high-precision map data segments corresponding to the high-precision map data; wherein, the multiple test perception data segments, the multiple test vehicle-road cooperative data segments, and the multiple high-precision map data segments correspond one-to-one.

[0036] It is also used to perform fusion processing on the corresponding test perception data segment, the test vehicle-road cooperative data segment and the high-precision map data segment to determine the scene restoration data segment;

[0037] It is also used to determine the scene restoration data of the target scene based on multiple scene restoration data segments.

[0038] To achieve the above objectives, in a fourth aspect, this application also provides a driving prediction device, the driving prediction device comprising:

[0039] The second acquisition module is used to acquire the original scene reconstruction data of the target scene; wherein the original scene reconstruction data is determined by the method of constructing scene reconstruction data as described in any one of the first aspects;

[0040] The driving prediction module is used to predict the driving scenario in the target scenario based on the original scene reconstruction data.

[0041] To achieve the above objectives, in a fifth aspect, this application also provides a vehicle, including: a processor and a memory, the processor being configured to execute a control program stored in the memory to achieve:

[0042] The method for constructing scene reconstruction data as described in the first aspect above; or...

[0043] The driving prediction method described in the second aspect above.

[0044] To achieve the above objectives, in a sixth aspect, this application also provides a storage medium storing one or more programs, which can be executed by one or more processors to perform:

[0045] The method for constructing scene reconstruction data as described in the first aspect above; or...

[0046] The driving prediction method described in the second aspect above.

[0047] The beneficial effects of this application are:

[0048] In this application, high-precision map data can be verified based on test perception data and test vehicle-road cooperative data of the target scene, thereby obtaining scene reconstruction data after the fusion of the three. This allows the scene reconstruction data to have the advantages of the above three types of data at the same time, so as to avoid the defect of insufficient data richness when using high-precision map data independently. This makes the scene reconstruction data more reliable, richer, and covers a wider area, improving the data stability, accuracy, and predictability in the reconstructed scene, and enhancing the user experience of scene reconstruction. Attached Figure Description

[0049] Figure 1 This illustration shows a flowchart of a method for constructing scene reconstruction data according to an embodiment of this application;

[0050] Figure 1a This diagram illustrates a data structure provided in an embodiment of this application.

[0051] Figure 2 This illustration shows a flowchart of a method for constructing scene reconstruction data according to an embodiment of this application;

[0052] Figure 2a This diagram illustrates an image corresponding to each data segment provided in an embodiment of this application.

[0053] Figure 2b This diagram illustrates an image corresponding to each data segment provided in an embodiment of this application.

[0054] Figure 2c This diagram illustrates an image corresponding to each data segment provided in an embodiment of this application.

[0055] Figure 3 This diagram illustrates a flow chart of a driving prediction method provided in an embodiment of this application.

[0056] Figure 4 This diagram illustrates the structure of an apparatus for constructing scene reconstruction data according to an embodiment of this application.

[0057] Figure 5 This diagram illustrates the structure of a driving prediction device according to an embodiment of this application.

[0058] Figure 6 This illustration shows a structural diagram of a vehicle according to an embodiment of this application;

[0059] in:

[0060] 10a, First Acquisition Module; 20a, Determination Module;

[0061] 10b, Second Acquisition Module; 20b, Driving Prediction Module;

[0062] 100. Vehicle; 101. Processor; 102. Memory; 1021. Operating system; 1022. Application program; 103. User interface; 104. Network interface; 105. Bus system. Detailed Implementation

[0063] The embodiments of this application will be described below with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. This application can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be understood that the preferred embodiments are only for illustrating this application and are not intended to limit the scope of protection of this application.

[0064] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of this application. Therefore, the drawings only show the components related to this application and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0065] The embodiments of this application will be described below with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. This application can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be understood that the preferred embodiments are only for illustrating this application and are not intended to limit the scope of protection of this application.

[0066] To facilitate understanding of the embodiments of this application, the following will provide further explanation and description with reference to the accompanying drawings and specific embodiments. These embodiments do not constitute a limitation on the embodiments of this application.

[0067] This embodiment provides a method for constructing scene reconstruction data, which can be applied to vehicles, including test vehicles. (Reference) Figure 1 As shown, the method may include:

[0068] S110. Acquire test perception data, test vehicle-road cooperative data and high-precision map data for the target scenario;

[0069] S120. The test perception data, test vehicle-road cooperative data, and high-precision map data are divided according to the same rules to determine multiple test perception data segments corresponding to the test perception data, multiple test vehicle-road cooperative data segments corresponding to the test vehicle-road cooperative data, and multiple high-precision map data segments corresponding to the high-precision map data; wherein, the multiple test perception data segments, the multiple test vehicle-road cooperative data segments, and the multiple high-precision map data segments correspond one-to-one.

[0070] S130. The corresponding test perception data segment, test vehicle-road cooperative data segment and high-precision map data segment are fused to determine the scene restoration data segment.

[0071] S140. Based on multiple scene restoration data segments, determine the scene restoration data for the target scene.

[0072] In step S110, the target scene can refer to the area where driving prediction needs to be performed based on scene reconstruction. It can be an area corresponding to a journey, an area corresponding to a city, or other areas, without limitation.

[0073] Test perception data refers to data collected by vehicle sensors. For example, test perception data may include image data collected by camera devices, detection data collected by radar, data collected by infrared sensors, etc., without limitation.

[0074] The test vehicle-to-infrastructure (V2X) data refers to data obtained based on V2X. V2X enables real-time dynamic interaction and linkage between vehicles, roads, people, and the environment. For example, various sensors, including cameras, millimeter-wave radar, and lidar, are directly installed on streetlight poles. The information collected by these sensors is then integrated and transmitted to the vehicle network and to each vehicle. The sensors on streetlight poles are more numerous and comprehensive, and their higher field of view and less obstruction result in better information collected than that from sensors on vehicles.

[0075] It should be noted that for environments with vehicle-road cooperative systems, more specific environmental information can be obtained, such as dynamic traffic environment participation information from traffic lights and electronic eyes that have been integrated with vehicle-road cooperative systems. This information is independent of the road dataset obtained from sensors on vehicles, which can improve the richness and reliability of the data.

[0076] High-definition (HD) maps are high-precision maps used for autonomous driving, containing map elements such as road shapes, road markings, traffic signs, and obstacles. Map accuracy can be down to the centimeter level.

[0077] In this step, the test vehicle can collect test perception data based on its various sensors and acquire test vehicle-to-infrastructure (V2I) data based on vehicle-to-infrastructure (V2I) technology. Additionally, the test vehicle can also acquire high-precision map data from the cloud or other devices via network communication. In other words, in this step, the test vehicle can simultaneously acquire test perception data, V2I data, and high-precision map data for the target scene, enabling it to obtain richer and more reliable data about the target scene.

[0078] In step S120, after the vehicle obtains the test perception data, test vehicle-road cooperative data and high-precision map data of the target scene, it can segment the above data according to the same rules to obtain multiple data segments with the same data type.

[0079] Specifically, the test perception data, the test vehicle-to-infrastructure (V2I) data, and the high-precision map data can be divided into multiple test perception data segments, V2I data segments, and high-precision map data segments. Furthermore, these multiple test perception data segments, V2I data segments, and high-precision map data segments correspond one-to-one. For example, after the test perception data, V2I data, and high-precision map data are divided, the corresponding test perception data segments, V2I data segments, and high-precision map data segments can correspond to the same area of ​​the target scene.

[0080] In this step, the location area corresponding to the target scene can be divided into multiple sub-regions with a set size. Then, the test perception data is segmented into multiple test perception data segments corresponding to each of the sub-regions. Similarly, the test vehicle-road cooperative data can be segmented into multiple test vehicle-road cooperative data segments corresponding to each of the sub-regions. Furthermore, the high-precision map data can be segmented into multiple high-precision map data segments corresponding to each of the sub-regions. The test perception data segment, test vehicle-road cooperative data segment, and high-precision map data segment corresponding to the same sub-region are mutually corresponding test perception data segments, test vehicle-road cooperative data segments, and high-precision map data segments.

[0081] The dimensions can be set according to the actual situation, and their specific values ​​are not limited. Understandably, different dimensions can correspond to sub-regions of different shapes and sizes. When dividing the location area corresponding to the target scene, the location (e.g., latitude and longitude) can be determined based on the navigation map data stored in the scene reconstruction device.

[0082] In some implementations...

[0083] The size can be set to 2km*100m, meaning that in this embodiment, the target scene can be divided into multiple square areas of 2km*100m. In this embodiment, a coordinate system can be established using latitude and longitude, thereby dividing the location area corresponding to the target scene into multiple square areas of 2km*100m based on latitude and longitude.

[0084] In this implementation, the test vehicle can first be injected with offline high-precision map data and navigation map data to ensure that the vehicle can still access the data in the scene reconstruction device even when driving in areas without network signal coverage. In the initial state, if the vehicle is connected to the network, the scene reconstruction device can automatically update the local high-precision map data and navigation map data from the cloud server to ensure data accuracy. In the learning state, that is, during the test vehicle's navigation and driving using the raw data from the scene reconstruction device, various sensors in the vehicle can collect data on the road sections the vehicle passes through, and the collected data can be saved as images, for example, by collecting image data at a frequency of 16 frames per second. Simultaneously, the collected 2D image data can be converted into 3D scene data, and then into planar test perception data.

[0085] It should be noted that the test perception data parsed in this embodiment, through the above method, can meet the data density requirements of most scenarios in most cases. After obtaining the test perception data, the vehicle can combine this data with the specific location in the target scene and save it. This location can be determined based on navigation map data and can be represented by latitude and longitude, or by other information; there is no limitation on this. Assuming that latitude and longitude are used in this embodiment.

[0086] In this embodiment, a location can be set every 2km along the longitude and every 100m along the latitude, that is, every 2km*100m is divided into an area, and the test sensing data of this area is taken as a test sensing data segment. This test sensing data segment can be marked with the midpoint, start point, end point, or other points of the corresponding area; there are no limitations on this. In this embodiment, the test sensing data segment can be marked with the latitude and longitude of the midpoint of the left wide side of the corresponding square area.

[0087] In addition to the perception data used for testing, more specific environmental information can be obtained for environments with test vehicle-road cooperative data, such as dynamic traffic environment participation information from traffic lights and electronic eyes that have already been integrated with vehicle-road cooperation. This information is independent of the data acquired by the sensors on the vehicle. This implementation method can also segment the test vehicle-road cooperative data acquired by the vehicle. Specifically, a location can be set every 2km along the longitude and every 100m along the latitude, that is, every 2km * 100m is divided into a region, and the test vehicle-road cooperative data of this region is considered as a test vehicle-road cooperative data segment.

[0088] In addition, vehicles can also segment high-precision map data. First, the original structure of the high-precision map data is decoupled. Then, it is segmented using the same segmentation rules as the test perception data and test vehicle-road cooperative data. Specifically, a location can be set every 2km along the longitude and every 100m along the latitude, meaning each 2km * 100m area is divided into a region, and the high-precision map data of that region is considered a single high-precision map data segment.

[0089] In other words, in this implementation, the final state of the test perception data, test vehicle-road cooperative data, and high-precision map data should be a 2km*100m square area as a data segment, and each data segment can be marked with a corresponding latitude and longitude point, that is, the latitude and longitude of the midpoint of the left wide side of the corresponding square area. These latitude and longitude points can be recorded as the starting latitude and longitude of the corresponding square area or the corresponding data segment. For more information on the data structure composed of latitude and longitude and data segments, please refer to [reference needed]. Figure 1a As shown, Figure 1a It may include n starting latitude and longitude coordinates and their corresponding n data segments (which are test perception data segments in test perception data, test vehicle-road cooperative data segments in test vehicle-road cooperative data, and high-precision map data segments in high-precision map data).

[0090] It should be noted that, in addition to segmenting the test perception data, test vehicle-road cooperative data, and high-precision map data to obtain corresponding test perception data segments, test vehicle-road cooperative data segments, and high-precision map data segments in the above manner, other methods can also be used to obtain them, and there are no limitations on this.

[0091] In step S130, after obtaining multiple sets of corresponding test perception data segments, test vehicle-road cooperative data segments, and high-precision map data segments through test measurements, the test perception data segments and test vehicle-road cooperative data segments in each set of corresponding test perception data segments, test vehicle-road cooperative data segments, and high-precision map data segments can be used to verify the high-precision map data segments. That is, the test perception data segments, test vehicle-road cooperative data segments, and high-precision map data segments in each set of corresponding test perception data segments, test vehicle-road cooperative data segments, and high-precision map data segments are fused to obtain the corresponding scene reconstruction data segments.

[0092] For example, after obtaining the data segments corresponding to each sub-region, the test perception data segment, test vehicle-road cooperative data segment, and high-precision map data segment corresponding to the same sub-region can be fused to obtain the scene reconstruction data segment corresponding to that sub-region. This scene reconstruction data segment has better data stability, data accuracy, and data predictability.

[0093] In step S140, after the test vehicle obtains scene reconstruction data segments corresponding to multiple sub-regions, it can construct complete scene reconstruction data of the target scene consisting of multiple sub-regions based on these scene reconstruction data segments. Scene reconstruction data has better data stability, data accuracy, and data predictability. Based on this scene reconstruction data, the target scene can be better reconstructed, which facilitates subsequent driving prediction based on this scene reconstruction data and improves the user experience of driving prediction.

[0094] In this method, high-precision map data can be verified based on test perception data and test vehicle-road cooperative data of the target scene, thereby obtaining scene reconstruction data after the fusion of the three. This allows the scene reconstruction data to have the advantages of the above three types of data at the same time, so as to avoid the defect of insufficient data richness when using high-precision map data independently. This makes the scene reconstruction data more reliable, richer, and covers a wider area, improving the data stability, accuracy, and predictability in the reconstructed scene, and enhancing the user experience of scene reconstruction.

[0095] This embodiment provides a method for constructing scene reconstruction data, which can be applied to vehicles, such as test vehicles. (Reference) Figure 2 As shown, in this method, the corresponding test perception data segment, test vehicle-road cooperative data segment, and high-precision map data segment are fused to determine the scene reconstruction data segment, which may include:

[0096] S210. The data in the test perception data segment, the test vehicle-road cooperative data segment and the high-precision map data segment are respectively processed by imaging in the form of data points to determine the perception image corresponding to the test perception data segment, the vehicle-road cooperative image corresponding to the test vehicle-road cooperative data segment and the high-precision map image corresponding to the high-precision map data segment.

[0097] S220. Perform difference analysis on perceived images, vehicle-road cooperative images, and high-precision map images;

[0098] S230. Based on the difference analysis results, select the scene restoration data segment from the test perception data segment, the test vehicle-road cooperative data segment, and the high-precision map data segment.

[0099] In step S210, after the test vehicle obtains the test perception data segment, test vehicle-road cooperative data segment and high-precision map data segment corresponding to each sub-region, it can select a set type of data from each data segment to mark points, so that the set type of data in each data segment is imaged in the form of data points, thereby obtaining the image corresponding to each data segment.

[0100] The data type settings can be configured according to actual needs, and there are no restrictions on the specific data types. For example, data type settings can include lane line data, traffic light data, traffic sign data, obstacle data, etc., without any limitations.

[0101] In some implementations...

[0102] refer to Figure 2a , Figure 2b or Figure 2c As shown, the data type can include lane line data. In this embodiment, lane line data in each test perception data segment can be plotted using black and white imaging to obtain the corresponding perception image for the test perception data segment. When plotting, the plotting positions corresponding to the lane line data can be determined based on latitude and longitude. Black dots in the perception image indicate that a lane line exists at the corresponding location. Similarly, vehicle-road cooperative images corresponding to each test vehicle-road cooperative data segment and high-precision map images corresponding to each high-precision map data segment can be obtained.

[0103] It should be noted that, in addition to the imaging processing described above, other methods can also be used for imaging processing, and there are no limitations on these methods. For example, when performing imaging processing, the data type can include multiple data types, and each data type can be imaged using data points of different colors.

[0104] In step S220, after the test vehicle obtains the perception image, vehicle-road cooperative image, and high-precision map image, it can perform difference analysis on the corresponding perception image, test vehicle-road cooperative data, and high-precision map image. That is, it can perform difference analysis on the perception image, vehicle-road cooperative image, and high-precision map image corresponding to the same sub-region to obtain the difference analysis results.

[0105] The difference analysis results can include identical regions with the same data points in corresponding perception images, test vehicle-road cooperative data, and high-precision map images, as well as different regions with different data points. Different regions can be further categorized into single regions where only one image contains data points, and difference regions where at least two images contain data points with discrepancies.

[0106] In step S230, after the test vehicle obtains the difference analysis results of the corresponding perception images, test vehicle-road cooperative data and high-precision map images, it can select corresponding data from the test perception data segment corresponding to the perception images, the test vehicle-road cooperative data segment corresponding to the test vehicle-road cooperative data and the high-precision map data segment corresponding to the high-precision map images based on the difference analysis results, thereby forming the final scene restoration data segment.

[0107] Specifically, for a first region where there are identical data points in the corresponding perception image, vehicle-road cooperative image, and high-precision map image, the first region scene restoration data corresponding to the first region can be selected from any data segment among the test perception data segment corresponding to the aforementioned perception image, the test vehicle-road cooperative data segment corresponding to the aforementioned test vehicle-road cooperative data, and the high-precision map data segment corresponding to the aforementioned high-precision map image.

[0108] For example, the perception image, vehicle-road cooperative image, and high-precision map image corresponding to a single sub-region may include, for example: Figure 2a The image corresponding to the first region shown. For the aforementioned first region, scene reconstruction data corresponding to the first region can be selected from any one of the following data segments: the test perception data segment corresponding to the aforementioned perception image, the test vehicle-road cooperative data segment corresponding to the aforementioned test vehicle-road cooperative data, and the high-precision map data segment corresponding to the aforementioned high-precision map image. This scene reconstruction data of the first region can then be used as part of the scene reconstruction data segment corresponding to the aforementioned single sub-region.

[0109] Specifically, for the second region where data points exist in one of the corresponding perception images, vehicle-road cooperative images, and high-precision map images, the second region scene restoration data can be selected from the data segment corresponding to the image where data points exist in the second region, which is the test perception data segment corresponding to the aforementioned perception image, the test vehicle-road cooperative data segment corresponding to the aforementioned test vehicle-road cooperative data, and the high-precision map data segment corresponding to the aforementioned high-precision map image.

[0110] For example, when a second region exists only in the vehicle-to-infrastructure (V2I) image, which is part of the perception image, vehicle-to-infrastructure (V2I) image, and high-precision map image corresponding to a single sub-region, data corresponding to this second region can be selected from the test V2I data segment corresponding to the V2I image as the scene reconstruction data for that second region. This second region scene reconstruction data can then be used as part of the scene reconstruction data segment corresponding to the single sub-region.

[0111] Among them, for the corresponding perception images, vehicle-road cooperative images and high-precision map images, if two images have data points and the data points are the same, the third region scene restoration data corresponding to the third region is selected from the data segment corresponding to either of the two images.

[0112] For example, when a third region exists in the perception image, vehicle-to-infrastructure (V2I) image, and high-precision map image corresponding to a single sub-region, and the data points differ only in the perception image and V2I image, data corresponding to the third region can be selected from either the test perception data segment corresponding to the perception image or the test V2I data segment corresponding to the V2I image (e.g., the test perception data segment or the test V2I data segment) as the third region scene reconstruction data. This third region scene reconstruction data can then be used as part of the scene reconstruction data segment corresponding to the single sub-region.

[0113] Among them, for the fourth region in the corresponding perception image, vehicle-road cooperative image and high-precision map image, at least two images have data points and the data points are different, the fourth region scene restoration data corresponding to the fourth region can be selected from the data segments with higher priority from the data segments corresponding to at least two images.

[0114] Among these, the priority of the test perception data segment is higher than that of the test vehicle-road cooperative data segment, which in turn is higher than that of the high-precision map data segment. In other words, when corresponding data for the fourth region exists in both the test perception data segment and the test vehicle-road cooperative data segment (or the high-precision map data segment), the data corresponding to the differing regions in the test perception data segment can be selected as the scene reconstruction data for the fourth region. Similarly, when corresponding data for the fourth region exists in both the test vehicle-road cooperative data segment and the high-precision map data segment, the data corresponding to the differing regions in the test vehicle-road cooperative data segment can be selected as the scene reconstruction data for the fourth region. Finally, when data for the fourth region exists in all three segments—the test perception data segment, the test vehicle-road cooperative data segment, and the high-precision map data segment—the data corresponding to the fourth region in the test perception data segment can be selected as the scene reconstruction data for the fourth region.

[0115] It should be noted that the priority of various data segments is not limited to the above settings; it can be set according to actual needs, and there are no restrictions on this.

[0116] In some implementations...

[0117] The perception image, vehicle-road cooperative image, and high-precision map image corresponding to a single sub-region may include, for example: Figure 2b The image corresponding to the fourth region shown. (Reference) Figure 2b It can be seen that the images in the fourth region are the same for both the perceived image and the vehicle-road cooperative image, while the image in the fourth region for the high-precision map is different from the two images mentioned above. In this embodiment, the priority of the test perception data segment is higher than that of the test vehicle-road cooperative data segment, and the priority of the test vehicle-road cooperative data segment is higher than that of the high-precision map data segment. Therefore, this embodiment can select the data corresponding to the fourth region from the test perception data segment corresponding to the perceived image as the scene reconstruction data for the fourth region.

[0118] In some implementations...

[0119] The perception image, vehicle-road cooperative image, and high-precision map image corresponding to a single sub-region may include, for example: Figure 2c The image corresponding to the fourth region shown. (Reference) Figure 2c It can be seen that the images in the fourth region are different from those in the perceived image, the vehicle-road cooperative image, and the high-precision map image. In this embodiment, the priority of the test perceived data segment is higher than that of the test vehicle-road cooperative data segment, and the priority of the test vehicle-road cooperative data segment is higher than that of the high-precision map data segment. Therefore, this embodiment can select the data corresponding to the fourth region from the test perceived data segment corresponding to the perceived image as the scene reconstruction data of the fourth region.

[0120] It should be noted that, in addition to selecting the scene restoration data of the fourth region corresponding to the fourth region through the above method, other methods can also be used, and there are no restrictions on this.

[0121] For example, in corresponding perception images, vehicle-road cooperative images, and high-precision map images, all three images contain data points. However, for areas where two of the images have the same data points, the scene reconstruction data corresponding to the aforementioned area can be selected from the data segments with higher priority among the data segments corresponding to the two images with the same data points.

[0122] Furthermore, in this step, after obtaining the scene reconstruction data corresponding to the various regions mentioned above, complete scene reconstruction data segments for the corresponding sub-regions can be constructed based on the scene reconstruction data of these multiple regions. After obtaining the scene reconstruction data segments corresponding to each sub-region, scene reconstruction data for the target scene can be constructed based on all the scene reconstruction data segments, thereby obtaining scene reconstruction data that can satisfy a journey, a city, or another specific region.

[0123] This method verifies high-precision map data based on test perception data and test vehicle-road cooperative data. It can take into account the problem of untimely updates of high-precision map data caused by zero-time road adjustments, thereby improving the data stability, data accuracy and data predictability in the restored scenario, and enhancing the user experience of scenario restoration.

[0124] This embodiment provides a driving prediction method applicable to vehicles, including passenger vehicles used by users. (Reference) Figure 3 As shown, the method may include:

[0125] S310. Obtain the original scene restoration data of the current scene;

[0126] S320: Based on the original scene reconstruction data, perform scene prediction for driving in the current scene.

[0127] In step S310, the original scene reconstruction data is determined by the method for constructing scene reconstruction data in the above embodiments. Specifically, after the test vehicle completes the scene reconstruction data of the target scene based on the method for constructing scene reconstruction data, the passenger vehicle can obtain the scene reconstruction data from the test vehicle, and this scene reconstruction data can be used as the original scene reconstruction data.

[0128] For example, after a test vehicle completes the scene reconstruction data of a target scene based on a method for constructing scene reconstruction data, the scene reconstruction data can be stored in a scene reconstruction device. Relevant personnel can then install the scene reconstruction device from the test vehicle into a passenger vehicle. The installation method can be plug-in or other methods, without limitation. After the scene reconstruction device is installed in the passenger vehicle, the passenger vehicle can obtain the scene reconstruction data from the scene reconstruction device, and this scene reconstruction data can be recorded as the original scene reconstruction data.

[0129] It should be noted that, in addition to the methods mentioned above, passenger vehicles can also obtain original scene reconstruction data through other means, and there are no restrictions on this.

[0130] In step S320, after the passenger vehicle obtains the original scene reconstruction data of the target scene, when the passenger vehicle is driving in the target scene, it can perform scene prediction based on the original scene reconstruction data to achieve more reliable driving prediction.

[0131] Among them, passenger vehicles can predict the scene during driving based on the original scene reconstruction data, or they can combine the perception data and / or the vehicle-road cooperative data used for testing to predict the scene.

[0132] Passenger vehicles can collect real-time driving perception data of their driving environment through various sensors. For vehicles and scenarios with vehicle-to-infrastructure (V2I) cooperation, the vehicles can also acquire real-time V2I data. Then, based on the acquired real-time driving perception data and / or V2I data, as well as the original scene reconstruction data, the vehicles can construct target scene reconstruction data.

[0133] The method for passenger vehicles to acquire driving perception data and driving vehicle-to-infrastructure (V2I) data can refer to the method for test vehicles to acquire test perception data and test V2I data in the above embodiments, and will not be elaborated further.

[0134] In some implementations...

[0135] Passenger vehicles can construct target scene reconstruction data based on driving perception data and original scene reconstruction data. Specifically, during operation, passenger vehicles collect driving perception data in real time through sensors. Based on this data, the corresponding perception area can be determined. Then, the corresponding original scene reconstruction data for that area is determined from the original scene reconstruction data. Finally, the driving perception data can directly replace the original scene reconstruction data for that area, thus obtaining the final target scene reconstruction data.

[0136] In some implementations...

[0137] Passenger vehicles can construct target scene reconstruction data based on vehicle-to-infrastructure (V2I) data and original scene reconstruction data. Specifically, during operation, after obtaining V2I data through V2I, the passenger vehicle can determine its corresponding region. Then, it can identify the corresponding region's original scene reconstruction data from the original scene reconstruction data. Finally, the V2I data can directly replace the original regional scene reconstruction data to obtain the final target scene reconstruction data.

[0138] In some implementations...

[0139] Passenger vehicles can construct target scene reconstruction data based on driving perception data, vehicle-to-infrastructure (V2I) data, and original scene reconstruction data. Specifically, during operation, passenger vehicles acquire driving perception data in real time through sensors, determine the corresponding perception area based on this data, and then extract the corresponding area's original scene reconstruction data from the original scene reconstruction data. Furthermore, during operation, passenger vehicles obtain driving V2I data through V2I, and can also retrieve the corresponding area's driving V2I data from this driving V2I data.

[0140] Then, the required type of data can be selected from the aforementioned vehicle perception data and plotted as points, so that the specified type of data in the vehicle perception data is imaged as data points, thereby obtaining the driving perception image corresponding to the driving perception data. The regional driving vehicle-road cooperative data and the regional original scene reconstruction data are processed in the same way to obtain the regional driving vehicle-road cooperative image corresponding to the regional driving vehicle-road cooperative data and the regional original scene reconstruction image corresponding to the regional original scene reconstruction data.

[0141] Then, a difference analysis is performed on the aforementioned driving perception images, regional driving vehicle-road cooperative images, and regional original scene reconstruction images. Based on the difference analysis results, regional target scene reconstruction data is selected from the driving perception data, regional driving vehicle-road cooperative data, and regional original scene reconstruction data. For specific implementation, refer to steps S220 and S230 in other embodiments, where the driving perception data corresponds to the test perception data segment, the regional driving vehicle-road cooperative data corresponds to the vehicle-road cooperative data segment, and the regional original scene reconstruction data corresponds to the high-precision map data segment.

[0142] It should be noted that, in addition to obtaining the target scene reconstruction data through the methods described above, other methods can also be used to obtain the target scene reconstruction data, and there are no limitations on these methods.

[0143] In this step, after the passenger vehicle obtains the target scene reconstruction data, it can store it and make scene predictions for driving in the target scene based on the target scene reconstruction data, thereby achieving more reliable driving prediction.

[0144] It should be noted that as the frequency of use of the scene reconstruction device increases, the amount of data accumulated in the device will also gradually increase, expanding its applicable area. Ultimately, this will enable scene reconstruction to achieve the same effect as scenarios with both high-precision map data and vehicle-to-infrastructure (V2I) data, even in situations lacking both. Simultaneously, it can continuously predict vehicle movement in areas undetectable by sensors, providing a solid data foundation for fully automated intelligent driving. In practical use, when the perception data is clear, the original scene reconstruction data can be re-verified in real time by combining the perception data collected by the vehicle and / or the V2I data acquired by the vehicle. This avoids erroneous predictions caused by temporary road changes or untimely data updates.

[0145] In this method, test vehicles can combine geographic information from navigation map data to label and record original scene reconstruction data. Simultaneously, an independent indexing system can be established within the data to organize it (e.g., based on multiple data segments), thus accelerating data retrieval when using the data. Furthermore, in different regions, corresponding data can be pre-filtered and constructed using different navigation map data to obtain scene reconstruction data for that region. When passenger vehicles travel in areas with scene reconstruction data, detailed lane, road surface, and surrounding environmental information can be retrieved based on the scene reconstruction data to predict vehicle control decisions. Furthermore, real-time driving perception data and / or vehicle-to-infrastructure (V2I) data can be combined to validate the scene reconstruction data, resulting in more reliable target scene reconstruction data and further improving the reliability of driving predictions.

[0146] This embodiment provides an apparatus for constructing scene reconstruction data, applicable to vehicles, including test vehicles. This apparatus can be used to implement the method for constructing scene reconstruction data described above. For example, refer to... Figure 4 As shown, the device may include:

[0147] The first acquisition module 10a is used to acquire test perception data, test vehicle-road cooperative data and high-precision map data of the target scene;

[0148] The determination module 20a is used to segment the test perception data, test vehicle-road cooperative data, and high-precision map data according to the same rules, so as to determine multiple test perception data segments corresponding to the test perception data, multiple test vehicle-road cooperative data segments corresponding to the test vehicle-road cooperative data, and multiple high-precision map data segments corresponding to the high-precision map data; wherein, the multiple test perception data segments, the multiple test vehicle-road cooperative data segments, and the multiple high-precision map data segments correspond one-to-one.

[0149] It is also used to fuse corresponding test perception data segments, test vehicle-road cooperative data segments, and high-precision map data segments to determine the scene reconstruction data segments;

[0150] It is also used to restore data segments from multiple scenarios and determine the scene restoration data for the target scenario.

[0151] This embodiment provides an apparatus for constructing scene reconstruction data, applicable to vehicles, including passenger vehicles. This apparatus can be used to implement the driving prediction method described above. For example, refer to... Figure 5 As shown, the device may include:

[0152] The second acquisition module 10b is used to acquire the original scene reconstruction data of the target scene;

[0153] The driving prediction module 20b is used to predict the driving scenario in the target scenario based on the original scene reconstruction data.

[0154] The original scene reconstruction data can be determined by the method for constructing scene reconstruction data as described in the above embodiments.

[0155] This embodiment provides an apparatus for constructing scene reconstruction data, which can be applied to vehicles, including passenger vehicles. (Reference) Figure 5 As shown, in this device,

[0156] The second acquisition module 10b can be used to acquire driving perception data and / or driving vehicle-road cooperative data;

[0157] The driving prediction module 20b constructs target scene reconstruction data based on driving perception data and / or driving vehicle-road cooperative data, as well as original scene reconstruction data;

[0158] It is also used to perform scene prediction for driving in the target scene based on the target scene reconstruction data.

[0159] This embodiment provides a vehicle. The vehicle can be a pure electric vehicle, a hybrid electric vehicle, or any other vehicle including an electric motor; there is no limitation in this regard. The vehicle can be used as the aforementioned test vehicle or as the aforementioned vehicle-use vehicle; there is no limitation in this regard either.

[0160] refer to Figure 6 As shown, the vehicle 100 includes at least one processor 101, a memory 102, at least one network interface 104, and other user interfaces 103. The various components in the vehicle 100 are coupled together via a bus system 105. It is understood that the bus system 105 is used to enable communication between these components. In addition to a data bus, the bus system 105 also includes a power bus, a control bus, and a status signal bus. However, for clarity, all buses are referred to as bus system 105.

[0161] The user interface 103 may include a display, keyboard, or click vehicle (e.g., mouse, trackball, touchpad, or touchscreen).

[0162] It is understood that the memory 102 in the embodiments of this application can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced Synchronous DRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), and Direct Rambus RAM (DRRAM). The memory 102 described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0163] In some implementations, memory 102 stores elements, executable units or data structures, or subsets thereof, or extended sets thereof: operating system 1021 and application program 1022.

[0164] The operating system 1021 includes various system programs, such as a framework layer, a core library layer, and a driver layer, used to implement various basic business functions and handle hardware-based tasks. The application program 1022 includes various applications, such as a media player and a browser, used to implement various application functions. Programs implementing the methods of this application embodiment can be included in the application program 1022.

[0165] In this embodiment of the application, the processor 101 executes the methods provided in each method embodiment by calling the program or instructions stored in the memory 102, specifically the program or instructions stored in the application program 1022.

[0166] The methods disclosed in the embodiments of this application can be applied to or implemented by the processor 101. The processor 101 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the integrated logic circuit of the hardware or by instructions in the form of software in the processor 101. The processor 101 may be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software units in the decoding processor. The software units may be located in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 102. Processor 101 reads the information in memory 102 and performs the above method in conjunction with its hardware.

[0167] It is understood that the embodiments described herein can be implemented in hardware, software, firmware, middleware, microcode, or a combination thereof. For hardware implementation, the processing unit can be implemented in one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing vehicles (DSPDs), programmable logic vehicles (PLDs), field-programmable gate arrays (FPGAs), general-purpose processors, controllers, microcontrollers, microprocessors, other electronic units for performing the functions described herein, or combinations thereof.

[0168] For software implementation, the techniques described herein can be implemented by units that perform the functions described herein. The software code can be stored in memory and executed by a processor. The memory can be implemented in the processor or external to the processor.

[0169] This application also provides a storage medium (computer-readable storage medium). This storage medium stores one or more programs. The storage medium may include volatile memory, such as random access memory; it may also include non-volatile memory, such as read-only memory, flash memory, hard disk, or solid-state drive; and it may also include combinations of the above types of memory.

[0170] When one or more programs in the storage medium can be executed by one or more processors to implement the above-described method of execution in the vehicle.

[0171] The processor is used to execute the vehicle control program stored in the memory to implement the above-described method of execution in the vehicle.

[0172] It should be noted that this storage medium can be used in the aforementioned test vehicle as well as in the aforementioned passenger vehicle, without limitation.

[0173] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0174] It should be noted that the terms "one implementation," "embodiment," "exemplary embodiment," and "some embodiments" used in the specification indicate that the described embodiment may include a specific feature, structure, or characteristic, but not every embodiment necessarily includes that specific feature, structure, or characteristic. Furthermore, such phrases do not necessarily refer to the same embodiment. Moreover, when a specific feature, structure, or characteristic is described in connection with an embodiment, implementing such a feature, structure, or characteristic in conjunction with other embodiments, whether explicitly described or not, is within the knowledge scope of those skilled in the art.

[0175] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or vehicle 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 vehicle. 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 vehicle that includes said element.

[0176] The above embodiments are merely preferred embodiments provided to fully illustrate this application, and the scope of protection of this application is not limited thereto. Equivalent substitutions or modifications made by those skilled in the art based on this application are all within the scope of protection of this application.

Claims

1. A method for constructing scene reconstruction data, characterized in that, The method for constructing scene reconstruction data includes: Acquire test perception data, test vehicle-road cooperative data, and high-precision map data for the target scenario; The test perception data, the test vehicle-road cooperative data, and the high-precision map data are segmented according to the same rules to determine multiple test perception data segments corresponding to the test perception data, multiple test vehicle-road cooperative data segments corresponding to the test vehicle-road cooperative data, and multiple high-precision map data segments corresponding to the high-precision map data; wherein, the multiple test perception data segments, the multiple test vehicle-road cooperative data segments, and the multiple high-precision map data segments correspond one-to-one. The corresponding test perception data segment, the test vehicle-road cooperative data segment, and the high-precision map data segment are fused to determine the scene reconstruction data segment; Based on multiple scene reconstruction data segments, determine the scene reconstruction data of the target scene; The process of fusing the corresponding test perception data segment, the test vehicle-road cooperative data segment, and the high-precision map data segment to determine the scene reconstruction data segment includes: The data in the test perception data segment, the test vehicle-road cooperative data segment, and the high-precision map data segment are respectively processed into images in the form of data points to determine the perception image corresponding to the test perception data segment, the vehicle-road cooperative image corresponding to the test vehicle-road cooperative data segment, and the high-precision map image corresponding to the high-precision map data segment. Perform difference analysis on the perceived image, the vehicle-road cooperative image, and the high-precision map image; Based on the difference analysis results, the scene restoration data segment is selected from the test perception data segment, the test vehicle-road cooperative data segment, and the high-precision map data segment; The selection of the scene reconstruction data segment based on the difference analysis results from the test perception data segment, the test vehicle-road cooperative data segment, and the high-precision map data segment includes: For a first region where there are identical data points among the perceived image, the vehicle-road cooperative image, and the high-precision map image, select the scene restoration data of the first region corresponding to the first region from any one of the test perception data segment, the test vehicle-road cooperative data segment, and the high-precision map data segment; For the second region where data points exist in one of the perception images, the vehicle-road cooperative images, and the high-precision map images, the scene restoration data of the second region corresponding to the second region is selected from the data segment corresponding to the image where data points exist in the second region. For the third region in the perception image, the vehicle-road cooperative image and the high-precision map image, where two images have the same data points, the scene restoration data of the third region corresponding to the third region is selected from the data segment corresponding to either of the two images; For a fourth region in which at least two of the perception images, the vehicle-road cooperative images, and the high-precision map images have data points and the data points are different, the scene restoration data of the fourth region corresponding to the fourth region is selected from the data segments with higher priority among the data segments corresponding to the at least two images. Among them, the priority of the test perception data segment is higher than the priority of the test vehicle-road cooperative data segment, and the priority of the test vehicle-road cooperative data segment is higher than the priority of the high-precision map data segment; The step of segmenting the test perception data, the test vehicle-road cooperative data, and the high-precision map data according to the same rules to determine multiple test perception data segments corresponding to the test perception data, multiple test vehicle-road cooperative data segments corresponding to the test vehicle-road cooperative data, and multiple high-precision map data segments corresponding to the high-precision map data includes: A coordinate system is established using latitude and longitude, and the location area corresponding to the target scene is determined based on the navigation map data stored in the scene reconstruction device; The target scene is divided into multiple square sub-regions of a set size; wherein, the latitude and longitude of the midpoint of the left wide side of the corresponding square region are used as the starting latitude and longitude of the corresponding data segment. The test perception data, the test vehicle-road cooperative data, and the high-precision map data are divided into multiple sub-regions to determine multiple test perception data segments, multiple test vehicle-road cooperative data segments, and multiple high-precision map data segments.

2. The method for constructing scene reconstruction data according to claim 1, characterized in that, The sensor data used for testing was obtained through the following methods: Acquire image data of the target scene; The image data is converted to determine the sensory data for testing.

3. A driving prediction method, characterized in that, The driving prediction method includes: Obtain the original scene reconstruction data of the target scene; wherein, the original scene reconstruction data is determined by the method for constructing scene reconstruction data as described in claim 1 or 2; Based on the original scene reconstruction data, scene prediction is performed for driving in the target scene.

4. The driving prediction method according to claim 3, characterized in that, The step of performing scene prediction for driving in the target scene based on the original scene reconstruction data includes: Acquire driving perception data and / or driving vehicle-to-infrastructure (V2I) data; Based on the driving perception data and / or the driving vehicle-road cooperative data, as well as the original scene reconstruction data, the target scene reconstruction data is constructed. Based on the target scene reconstruction data, scene prediction is performed for driving in the target scene.

5. An apparatus for constructing scene reconstruction data, characterized in that, The apparatus for constructing scene reconstruction data includes: The first acquisition module is used to acquire test perception data, test vehicle-road cooperative data and high-precision map data of the target scene; The determination module is used to segment the test perception data, the test vehicle-road cooperative data, and the high-precision map data according to the same rules, so as to determine multiple test perception data segments corresponding to the test perception data, multiple test vehicle-road cooperative data segments corresponding to the test vehicle-road cooperative data, and multiple high-precision map data segments corresponding to the high-precision map data; wherein, the multiple test perception data segments, the multiple test vehicle-road cooperative data segments, and the multiple high-precision map data segments correspond one-to-one. It is also used to perform fusion processing on the corresponding test perception data segment, the test vehicle-road cooperative data segment and the high-precision map data segment to determine the scene restoration data segment; It is also used to determine the scene restoration data of the target scene based on multiple scene restoration data segments; The step of fusing the corresponding test perception data segment, the test vehicle-road cooperative data segment, and the high-precision map data segment to determine the scene reconstruction data segment includes: The data in the test perception data segment, the test vehicle-road cooperative data segment, and the high-precision map data segment are respectively processed into images in the form of data points to determine the perception image corresponding to the test perception data segment, the vehicle-road cooperative image corresponding to the test vehicle-road cooperative data segment, and the high-precision map image corresponding to the high-precision map data segment. Perform difference analysis on the perceived image, the vehicle-road cooperative image, and the high-precision map image; Based on the difference analysis results, the scene restoration data segment is selected from the test perception data segment, the test vehicle-road cooperative data segment, and the high-precision map data segment; The selection of the scene reconstruction data segment based on the difference analysis results from the test perception data segment, the test vehicle-road cooperative data segment, and the high-precision map data segment includes: For a first region where there are identical data points among the perceived image, the vehicle-road cooperative image, and the high-precision map image, select the scene restoration data of the first region corresponding to the first region from any one of the test perception data segment, the test vehicle-road cooperative data segment, and the high-precision map data segment; For the second region where data points exist in one of the perception images, the vehicle-road cooperative images, and the high-precision map images, the scene restoration data of the second region corresponding to the second region is selected from the data segment corresponding to the image where data points exist in the second region. For the third region in the perception image, the vehicle-road cooperative image and the high-precision map image, where two images have the same data points, the scene restoration data of the third region corresponding to the third region is selected from the data segment corresponding to either of the two images; For a fourth region in which at least two of the perception images, the vehicle-road cooperative images, and the high-precision map images have data points and the data points are different, the scene restoration data of the fourth region corresponding to the fourth region is selected from the data segments with higher priority among the data segments corresponding to the at least two images. Among them, the priority of the test perception data segment is higher than the priority of the test vehicle-road cooperative data segment, and the priority of the test vehicle-road cooperative data segment is higher than the priority of the high-precision map data segment; The step of segmenting the test perception data, the test vehicle-road cooperative data, and the high-precision map data according to the same rules to determine multiple test perception data segments corresponding to the test perception data, multiple test vehicle-road cooperative data segments corresponding to the test vehicle-road cooperative data, and multiple high-precision map data segments corresponding to the high-precision map data includes: A coordinate system is established using latitude and longitude, and the location area corresponding to the target scene is determined based on the navigation map data stored in the scene reconstruction device; The target scene is divided into multiple square sub-regions of a set size; wherein, the latitude and longitude of the midpoint of the left wide side of the corresponding square region are used as the starting latitude and longitude of the corresponding data segment. The test perception data, the test vehicle-road cooperative data, and the high-precision map data are divided into multiple sub-regions to determine multiple test perception data segments, multiple test vehicle-road cooperative data segments, and multiple high-precision map data segments.

6. A driving prediction device, characterized in that, The driving prediction device includes: The second acquisition module is used to acquire the original scene reconstruction data of the target scene; wherein, the original scene reconstruction data is determined by the method for constructing scene reconstruction data as described in claim 1 or 2; The driving prediction module is used to predict the driving scenario in the target scenario based on the original scene reconstruction data.

7. A vehicle, characterized in that, include: A processor and a memory, the processor being configured to execute a control program stored in the memory to achieve: The method for constructing scene reconstruction data as described in claim 1 or 2; or, The driving prediction method as described in claim 3 or 4.

8. A storage medium, characterized in that, The storage medium stores one or more programs, which can be executed by one or more processors to achieve: The method for constructing scene reconstruction data as described in claim 1 or 2; or, The driving prediction method as described in claim 3 or 4.

Images