Road network topological map generation method, electronic equipment and storage medium

Abstract

The invention discloses a road network topological map generation method, electronic equipment and a storage medium. The method comprises the following steps: acquiring vehicle track data corresponding to a parking lot, wherein the parking lot comprises at least two floors; processing the vehicle trajectory data to obtain at least two two-dimensional grid charts corresponding to the at least two floors; determining a plurality of key nodes in each two-dimensional grid chart, generating link information corresponding to the plurality of key nodes, and determining link direction information corresponding to the link information; and generating the road network topology map according to the plurality of key nodes corresponding to each two-dimensional grid chart, the link information and the link direction information. The map construction efficiency of the parking lot is improved.

Description

Technical Field

[0001] The embodiments disclosed in this application relate to the field of autonomous driving technology, and more specifically, to a method for generating a road network topology map, an electronic device, and a storage medium. Background Technology

[0002] As parking space-to-parking functions become increasingly common in autonomous driving tasks, they encompass the entire process from parking garage exit to public road navigation, parking garage cruising, and parking space entry. However, in the construction of parking lot topology maps in multi-story parking garage scenarios, the heavy reliance on manual annotation leads to low map construction efficiency, and the intersection recognition model lacks generalization ability, making it difficult to adapt to various heterogeneous parking lots. Summary of the Invention

[0003] According to embodiments of this application, this application proposes a method for generating a road network topology map, an electronic device, and a storage medium to solve the above-mentioned problems.

[0004] The first aspect of this application discloses a method for generating a road network topology map, comprising: acquiring vehicle trajectory data corresponding to a parking lot, wherein the parking lot includes at least two floors; processing the vehicle trajectory data to obtain at least two two-dimensional grid maps corresponding to the at least two floors; determining multiple key nodes in each of the two-dimensional grid maps, generating link information corresponding to the multiple key nodes, and determining link direction information corresponding to the link information; and generating the road network topology map based on the multiple key nodes, the link information, and the link direction information corresponding to each of the two-dimensional grid maps.

[0005] In some embodiments, processing the vehicle trajectory data to obtain at least two two-dimensional grid maps corresponding to the at least two floors includes: dividing the vehicle trajectory data according to the at least two floors to obtain sub-trajectory data corresponding to each floor; downsampling the sub-trajectory data using a two-dimensional grid to obtain a two-dimensional grid map corresponding to the floor, wherein the number of trajectory points in each grid in the two-dimensional grid map is greater than or equal to a preset value, and each grid represents the average three-dimensional coordinates of the trajectory points within the grid.

[0006] In some embodiments, determining multiple key nodes in the two-dimensional mesh graph includes: performing dilation and closure operations on the original trajectory region in the two-dimensional mesh graph to obtain the trajectory region in the two-dimensional mesh graph; refining the trajectory region in the two-dimensional mesh graph to obtain the skeleton line mesh corresponding to the road centerline; and determining the multiple key nodes in the skeleton line mesh.

[0007] In some embodiments, the plurality of key nodes include a first type of node and a second type of node; In the skeleton grid, determining the plurality of key nodes includes: utilizing the neighborhood relationship of the skeleton grid, retaining a first skeleton grid point and a second skeleton grid point in the skeleton grid, wherein the number of neighborhood points of the first skeleton grid point in the skeleton grid is equal to a first preset value, and the number of neighborhood points of the second skeleton grid point in the skeleton grid is greater than a second preset value, wherein the second preset value is greater than the first preset value; using the first skeleton grid point as a first type node, and in response to the existence of a neighborhood connectivity greater than or equal to a third preset value within a preset-size window used at the second skeleton grid point in the skeleton grid, using the second skeleton grid point as a second type node; using a third skeleton grid point in the skeleton grid as a second type node, wherein the third skeleton grid point corresponds to the ramp connection between two adjacent floors.

[0008] In some embodiments, generating link information corresponding to the plurality of key nodes includes: determining a current search point, wherein the current search point is one of the multiple key nodes; in a skeleton line grid, recursively determining a skeleton line grid point located in the extension direction of the skeleton line grid from the neighborhood points of the current search point, as the next search point; in the recursion, the angle between the current link direction corresponding to the current search point and the direction of the next search point is less than a preset angle, wherein the current link direction represents the direction between the current search point and the previous key node; in response to the next search point being another key node among the multiple key nodes, generating link information between the current search point and the next search point, wherein the link information includes data of the current search point and the next search point.

[0009] In some embodiments, the vehicle trajectory data includes trip identifiers and timestamp information; determining the link direction information corresponding to the link information includes: classifying the vehicle trajectory data according to the trip identifiers to obtain multiple vehicle trip data; for each vehicle trip data, sorting the trajectory point sequence corresponding to the vehicle trip data according to the timestamp information to obtain the order of passing through the second type of nodes among the multiple key nodes; and using the order of passing through the second type of nodes among the multiple key nodes for the link direction information.

[0010] In some embodiments, the vehicle trajectory data includes trip identifiers and timestamp information; determining the link direction information corresponding to the link information includes: classifying the vehicle trajectory data according to the trip identifiers to obtain multiple vehicle trip data; for each vehicle trip data, sorting the trajectory point sequence corresponding to the vehicle trip data according to the timestamp information to obtain the order of passing through the second type of nodes among the multiple key nodes; in response to the proportion of the number of vehicle trip data corresponding to the order of passing through the second type of nodes among the multiple key nodes being greater than a preset value, using the order of passing through the second type of nodes among the multiple key nodes for the link direction information.

[0011] In some embodiments, the road network topology map is generated based on the plurality of key nodes, link information and link direction information corresponding to each of the two-dimensional grid maps, and further includes: connecting the key nodes located at the ramp connection among the plurality of key nodes corresponding to the two-dimensional grid maps of adjacent floors for use in the road network topology map.

[0012] The second aspect of this application discloses an electronic device including a memory and a processor coupled to each other, the processor being configured to execute program instructions stored in the memory to implement the road network topology map generation method described in the first aspect.

[0013] The third aspect of this application discloses a non-volatile computer-readable storage medium storing program instructions thereon, which, when executed by a processor, implement the method for generating a road network topology map as described in the first aspect.

[0014] The beneficial effects of this application are as follows: obtaining vehicle trajectory data corresponding to parking lots, wherein the parking lot includes at least two floors, processing the vehicle trajectory data to obtain at least two two-dimensional grid maps corresponding to at least two floors, determining multiple key nodes in each two-dimensional grid map, generating link information corresponding to multiple key nodes, and determining the link direction information corresponding to the link information, and then generating a road network topology map based on multiple key nodes, link information and link direction information corresponding to each two-dimensional grid map, which can adapt to diverse parking lot layouts and improve the efficiency of parking lot map construction. Attached Figure Description

[0015] The present application will be further described below with reference to the accompanying drawings and embodiments. In the drawings: Figure 1 This is a flowchart illustrating the method for generating a road network topology map according to an embodiment of this application; Figure 2 This is a schematic diagram of a two-dimensional mesh diagram according to an embodiment of this application; Figure 3This is a schematic diagram of a skeleton line mesh according to an embodiment of this application; Figure 4 This is a schematic diagram of a skeleton line mesh according to an embodiment of this application; Figure 5 This is a schematic diagram of the structure of an electronic device according to an embodiment of this application; Figure 6 This is a schematic diagram of the structure of a non-volatile computer-readable storage medium according to an embodiment of this application. Detailed Implementation

[0016] In this application, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0017] In this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. Additionally, the character " / " generally indicates that the preceding and following related objects are in an "or" relationship. Furthermore, "many" in this application means two or more. Moreover, the term "at least one" in this application means any combination of at least two of any one or more of a plurality of objects. For example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C. Furthermore, the terms "first," "second," and "third" in this application are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features.

[0018] To enable those skilled in the art to better understand the technical solution of this application, the technical solution of this application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0019] Please see Figure 1 , Figure 1 This is a flowchart illustrating a method for generating a road network topology map according to an embodiment of this application. The subject executing this method can be an electronic device with computing capabilities, such as a microcomputer, server, laptop, tablet, or other mobile device, and may also include in-vehicle equipment, such as autonomous vehicles.

[0020] It should be noted that if substantially the same result is obtained, the method of this application is not based on... Figure 1 The sequence of processes shown is limited.

[0021] In some possible implementations, this method can be implemented by the processor calling computer-readable instructions stored in memory, such as... Figure 1 As shown, the method may include the following steps: S11: Obtain vehicle trajectory data corresponding to the parking lot, wherein the parking lot includes at least two floors.

[0022] The system acquires vehicle trajectory data corresponding to the parking lot. This trajectory data can include the three-dimensional coordinates (X / Y / Z) of trajectory points. For example, it could be the trajectory points of n vehicle trajectories generated by m vehicles traveling within the parking lot. Vehicle trajectory data can be generated by fusing information from multiple sources of sensors, such as IMU (Inertial Measurement Unit), GNSS (Global Navigation Satellite System), wheel speedometers, lidar, and cameras, and processed using odometry. The parking lot comprises at least two floors, with at least two floors connected. Vehicle trajectories include those of vehicles traveling on any one floor, and those traveling between multiple floors.

[0023] S12: Process the vehicle trajectory data to obtain at least two two-dimensional grid maps corresponding to at least two floors.

[0024] The vehicle trajectory data is processed to obtain at least two two-dimensional grid maps corresponding to at least two floors. For example, the parking lot area is divided into a fine grid, and each trajectory point in the vehicle trajectory data falls into one grid. For example, a grid with a trajectory point is designated as 1, and an empty grid is designated as 0, thus obtaining the corresponding two-dimensional grid map. The vehicle trajectory data can include the trajectory points corresponding to n vehicle trajectories generated by m vehicles driving in the parking lot. In the two-dimensional grid map, each grid can also store specific information, such as occupancy probability and passage attributes. By using at least two two-dimensional grid maps corresponding to at least two floors, the drivable area, obstacle distribution, or road structure in the parking lot can be inferred.

[0025] S13: Identify multiple key nodes in each two-dimensional grid diagram, generate link information corresponding to multiple key nodes, and determine the link direction information corresponding to the link information.

[0026] Key nodes can be drivable road endpoints or intersections within a parking lot. Multiple key nodes are identified in each 2D grid diagram, and corresponding link information is generated for each key node. For example, by performing path searches on multiple key nodes in each 2D grid diagram, corresponding links, such as lanes and ramps, can be generated. Furthermore, the link direction information is determined, including unidirectional and bidirectional directions.

[0027] S14: Generate a road network topology map based on multiple key nodes, link information and link direction information corresponding to each two-dimensional grid map.

[0028] Based on the multiple key nodes, link information, and link direction information corresponding to each two-dimensional grid map, a road network topology map is generated. For example, by obtaining the endpoints, intersections, and their corresponding links and links, a road network topology map can be generated. The road network topology map includes intersections, endpoints, and corresponding links in the parking lot.

[0029] In this embodiment, vehicle trajectory data corresponding to the parking lot is obtained. The parking lot includes at least two floors. The vehicle trajectory data is processed to obtain at least two two-dimensional grid maps corresponding to the at least two floors. Multiple key nodes in each two-dimensional grid map are determined, link information corresponding to the multiple key nodes is generated, and link direction information corresponding to the link information is determined. Then, based on the multiple key nodes, link information, and link direction information corresponding to each two-dimensional grid map, a road network topology map can be generated, which can adapt to diverse parking lot layouts and improve the efficiency of parking lot map construction.

[0030] In some embodiments, the vehicle trajectory data is processed to obtain at least two two-dimensional grid maps corresponding to at least two floors, including: dividing the vehicle trajectory data according to at least two floors to obtain sub-trajectory data corresponding to each floor; downsampling the sub-trajectory data using a two-dimensional grid to obtain a two-dimensional grid map corresponding to the floor, wherein the number of trajectory points in each grid in the two-dimensional grid map is greater than or equal to a preset value, and each grid represents the average three-dimensional coordinates of the trajectory points within the grid.

[0031] Vehicle trajectory data is divided into at least two floors to obtain sub-trajectory data for each floor. For example, crowdsourced trajectory data of vehicles in an n-story parking lot can be obtained and divided by floor to obtain a set of sub-trajectories for each of the n floors. The sub-trajectory data is downsampled using a two-dimensional grid to obtain a two-dimensional grid map corresponding to each floor. For example, a two-dimensional grid is constructed for each floor, the grid resolution is set, the X and Y coordinates of the trajectory points are sampled, and the average three-dimensional coordinates of all trajectory points within the grid are recorded to facilitate the recovery of the Link height. That is, each grid can represent the average three-dimensional coordinates of the trajectory points within the grid.

[0032] In some examples, underground parking garages may have multiple levels, such as B1, B2, etc., allowing trajectory data to be processed separately for each level. For instance, a single-level parking garage might have an area of ​​100m × 100m. Setting a grid resolution (e.g., 0.2m × 0.2m) divides the floor plan into 500 × 500 grid cells. Each grid represents a small area, and after a trajectory point falls into its corresponding grid, its average three-dimensional coordinates (X, Y, Z average values) are retained. For example, if a grid contains 10 trajectory points with an average X coordinate of 10.6, an average Y coordinate of 6.8m, and an average Z coordinate of -2.0m (relative height), then the center point of that grid would be recorded as (10.6, 6.8, -2.0).

[0033] In one implementation, the preset value ranges from 4 to 8. For example, the preset value is 6. A two-dimensional grid is used to downsample the sub-trajectory data, identifying grids with a number of trajectory points greater than or equal to 6 as target grids, thereby obtaining a two-dimensional grid map corresponding to each floor. Figure 2 As shown, Figure 2 This is a schematic diagram of a two-dimensional grid diagram according to an embodiment of this application. In the diagram, gray grids represent the number of corresponding trajectory points that is greater than or equal to 6.

[0034] In this context, the number of trajectory points in each grid of the two-dimensional grid is greater than or equal to a preset value. For example, after constructing a two-dimensional grid by floor based on the number of vehicle trajectories, the number of trajectory points in each grid can be counted, and grid areas with a vote count lower than a set threshold can be eliminated, thereby reducing noise interference.

[0035] In some embodiments, determining multiple key nodes in a two-dimensional mesh graph includes: performing dilation and closure operations on the original trajectory region in the two-dimensional mesh graph to obtain a trajectory region in the two-dimensional mesh graph; refining the trajectory region in the two-dimensional mesh graph to obtain a skeleton line mesh corresponding to the road centerline; and determining multiple key nodes in the skeleton line mesh.

[0036] The original trajectory region in the 2D mesh image is dilated and closed to obtain the trajectory region in the 2D mesh image. For example, following the processing method of 2D binary images, dilation and closing operations are performed on the mesh to eliminate mesh holes. Here, a threshold can be set to divide the pixels of the grayscale image into two categories: those with grayscale values ​​greater than the threshold are set to white, and those less than the threshold are set to black, to reduce the amount of data while highlighting the outline of the vehicle trajectory or specific areas. The trajectory region in the 2D mesh image is then refined to obtain the skeleton line mesh corresponding to the road centerline, such as... Figure 3 As shown, Figure 3This is a schematic diagram of a skeleton line mesh according to an embodiment of this application. The skeleton line mesh can serve as a reference line in path planning, allowing vehicles to travel safely and smoothly along this centerline. For example, the skeleton line width is one pixel. For instance, the thinned skeleton line can be extracted using the Zhang-Suen-Thinning algorithm. The Zhang-Suen-Thinning algorithm refines the shape by iteratively deleting boundary points, preserving the topological features of the original shape while maintaining connectivity.

[0037] In some embodiments, the multiple key nodes include first type nodes and second type nodes; in the skeleton line grid, determining the multiple key nodes includes: utilizing the neighborhood relationship of the skeleton line grid, retaining first skeleton line grid points and second skeleton line grid points in the skeleton line grid, wherein the number of neighborhood points of the first skeleton line grid point in the skeleton line grid is equal to a first preset value, and the number of neighborhood points of the second skeleton line grid point in the skeleton line grid is greater than a second preset value, wherein the second preset value is greater than the first preset value; using the first skeleton line grid point as a first type node, and in response to the existence of a neighborhood connectivity greater than or equal to a third preset value within a preset size window used at the second skeleton line grid point in the skeleton line grid, using the second skeleton line grid point as a second type node; using a third skeleton line grid point in the skeleton line grid as a second type node, wherein the third skeleton line grid point corresponds to the ramp connection between two adjacent floors.

[0038] Utilizing the neighborhood relationships of the skeleton line mesh, the first and second skeleton line mesh points are retained. For example, using the 8-neighborhood of the skeleton line mesh, the first and second skeleton line mesh points are determined. Specifically, the number of neighboring points of the first skeleton line mesh point in the skeleton line mesh is equal to a first preset value, for example, the first preset value is 1, meaning it has 1 neighboring point, and can be identified as the first skeleton line mesh point. The number of neighboring points of the second skeleton line mesh point in the skeleton line mesh is greater than the second preset value, for example, the second preset value is 2, meaning it has more than 2 neighboring points, and can be identified as the second skeleton line mesh point.

[0039] Multiple key nodes include first-type nodes and second-type nodes. For example, first-type nodes are endpoints, and second-type nodes are intersection points. First-type nodes are used as first-type nodes; that is, points with a neighboring node of 1 in the 8-neighborhood of the skeleton line grid are endpoints. In response to the presence of a neighboring connectivity greater than or equal to a third preset value within a preset-size window at a second-type skeleton line grid point, the second-type node is used as a second-type node. In some examples, if there are more than 2 neighboring nodes in the 8-neighborhood of the skeleton line grid, it is a potential intersection point. A 3×3 local window can then be used at the potential intersection point location. Removing the center point from the 3×3 grid reveals several connected components. A four-neighborhood check can be used; if there are three or more connected components, it can be identified as an intersection point. Understandably, by pre-calculating all possible 3×3 neighborhood cases and storing the pixel deletion decision in a lookup table, the decision to delete a grid point is directly determined during computation by looking up the table, thereby optimizing the refinement speed of the trajectory region in the 2D mesh graph.

[0040] Furthermore, the third skeleton line grid point in the skeleton line grid can also be used as a second type of node, where the third skeleton line grid point corresponds to the ramp connection between two adjacent floors. For example, based on the ramp entry and exit points, the corresponding skeleton grid point at the level floor location where the ramp connects can be set as an intersection point.

[0041] In some examples, such as Figure 4 As shown, Figure 4 This is a schematic diagram of a skeleton line mesh according to an embodiment of this application. Grid point N1 has 1 neighboring point in its 8-neighborhood, meaning N1 is a first skeleton line mesh point and can be used as a first type of node. The 8-neighborhood of N1 includes the 8 mesh points surrounding N1. Grid point N2 has more than 2 neighboring points in its 8-neighborhood and a connectivity of 3, meaning N2 is a second skeleton line mesh point and can be used as a second type of node. Grid point N3 has more than 2 neighboring points in its 8-neighborhood and a connectivity of 4, meaning N3 is also a second skeleton line mesh point and can be used as a second type of node. Furthermore, mesh points with 0 neighboring points in their 8-neighborhood can be deleted.

[0042] In some embodiments, generating link information corresponding to multiple key nodes includes: determining a current search point, wherein the current search point is one of the multiple key nodes; in the skeleton line mesh, recursively determining skeleton line mesh points located in the extension direction of the skeleton line mesh from the neighborhood points of the current search point, as the next search point; in the recursion, the angle between the current link direction corresponding to the current search point and the direction of the next search point is less than a preset angle, wherein the current link direction represents the direction between the current search point and the previous key node; in response to the next search point being another key node among the multiple key nodes, generating link information between the current search point and the next search point, wherein the link information includes data of the current search point and the next search point.

[0043] The current search point is determined, where it is one of several key nodes, such as a node that has never been visited. Within the skeleton grid, points in the neighboring grid that extend along the skeleton grid's direction are recursively selected as the next search point. For example, a depth-first search (DFS) is initiated from the current search point, where each current search point is searched from its eight neighboring directions to ensure all routes at intersections are fully explored. During the recursion, the angle between the current link direction and the direction of the next search point is less than a preset angle, where the current link direction represents the direction between the current search point and the previous key node. In some examples, the preset angle is between 80 and 100 degrees. For instance, to find a suitable next skeleton point from the eight neighbors, the angle between the link direction obtained from the current DFS search and the direction of the next skeleton point must be less than 90 degrees. The cosine value can be calculated based on the direction vectors of the previous skeleton point, the current skeleton point, and the next skeleton point; a cosine value greater than 0 indicates a valid point. Furthermore, the points that can be depth-searched are sorted by cosine value, and the skeleton points that have been visited are marked in real time, and the points are extended in reasonable directions along the skeleton line grid. Marked grids are no longer searched.

[0044] In response to the next search point being another key node among multiple key nodes, link information between the current search point and the next search point is generated. The link information includes data between the current search point and the next search point. For example, if another node is encountered during the search process, such as the first type node and / or the second type node mentioned above, a Link is generated and the path search is terminated. The corresponding Link's attachment point (Node) is recorded.

[0045] In some embodiments, vehicle trajectory data includes trip identifiers and timestamp information; determining the link direction information corresponding to the link information includes: classifying the vehicle trajectory data according to the trip identifiers to obtain multiple vehicle trip data; for each vehicle trip data, sorting the trajectory point sequence corresponding to the vehicle trip data according to the timestamp information to obtain the order of passing through the second type of nodes among multiple key nodes; and using the order of passing through the second type of nodes among multiple key nodes for the link direction information.

[0046] Vehicle trajectory data includes trip identifiers and timestamp information. For example, trip identifiers can be information corresponding to the trajectory, such as trajectory 1, trajectory 2, ..., trajectory n. Each trajectory is assigned a corresponding identifier for classification and other operations. In some examples, vehicle trajectory data may include three-dimensional coordinates (X / Y / Z), timestamps, trip identifiers, and floor identifiers.

[0047] Vehicle trajectory data is categorized according to trip identifiers, resulting in multiple vehicle trip data sets. For example, the trajectories of different vehicles on the same floor can be grouped together based on trip identifiers. For each vehicle trip data set, the sequence of trajectory points corresponding to the vehicle trip data is sorted according to timestamp information to obtain the order of passing through secondary nodes among multiple key nodes. This order of passing through secondary nodes among multiple key nodes is used for link direction information. For example, by sorting the trajectory point sequence by timestamp, the order in which the trajectory points pass through secondary nodes (intersection points) determines whether the link is unidirectional or bidirectional. In some examples, if the trajectory travels from intersection 1 to intersection 2 and then back from intersection 2 to intersection 1 in chronological order, the link between intersection 1 and intersection 2 is considered bidirectional. If there is only a link from intersection 1 to intersection 2, the link between intersection 1 and intersection 2 is considered unidirectional.

[0048] In some embodiments, determining the link direction information corresponding to the link information includes: classifying vehicle trajectory data according to the trip identifier to obtain multiple vehicle trip data; for each vehicle trip data, sorting the trajectory point sequence corresponding to the vehicle trip data according to the timestamp information to obtain the order of passing through the second type of nodes among multiple key nodes; in response to the proportion of the number of vehicle trip data corresponding to the order of passing through the second type of nodes among multiple key nodes to the multiple vehicle trip data being greater than a preset value, using the order of passing through the second type of nodes among multiple key nodes for the link direction information.

[0049] Vehicle trajectory data is categorized according to trip identifiers, resulting in multiple vehicle trip data sets. For example, the trajectories of different vehicles on the same floor can be grouped into one category based on the trip identifier. For each vehicle trip data set, the sequence of trajectory points corresponding to the vehicle trip data is sorted according to timestamp information to obtain the order of passing through the second type of nodes among multiple key nodes. This order of passing through the second type of nodes among multiple key nodes can include traveling from intersection 1 to intersection 2, then from intersection 2 to intersection 1, or only from intersection 1 to intersection 2, or only from intersection 2 to intersection 1.

[0050] Furthermore, in response to a situation where the proportion of vehicle trip data corresponding to the order of passing through the second type of nodes among multiple key nodes exceeds a preset value, the order of passing through the second type of nodes among multiple key nodes is used for link direction information. For example, the final link order can be determined based on a voting mechanism. For instance, in the intersection link order obtained from multiple trips, if the percentage of trips from intersection 1 to intersection 2 that are bidirectional exceeds a set threshold, it is considered bidirectional; if the percentage of trips from intersection 1 to intersection 2 that are bidirectional exceeds a set threshold, it is considered unidirectional.

[0051] In some embodiments, a road network topology map is generated based on multiple key nodes, link information and link direction information corresponding to each two-dimensional grid map, and further includes: connecting the key nodes located at the ramp connection among the multiple key nodes corresponding to the two-dimensional grid maps of two adjacent floors for use in the road network topology map.

[0052] In one example, key nodes located at ramp connections are connected among multiple key nodes in the two-dimensional grid diagrams corresponding to adjacent floors to create a road network topology map. For example, complete topology information is obtained based on Link, endpoint, and intersection information. Specifically, based on the intersection points (Nodes) that record the corresponding Links, and based on the link direction information determined by the order of passing through the second type of nodes among multiple key nodes, the in-degree and out-degree Links of each Node can be obtained, thereby generating the topology relationship.

[0053] Understandably, this application employs a crowdsourced trajectory point statistics method based on a two-dimensional downsampling grid to eliminate low-vote areas and reduce noise interference. The two-dimensional downsampling grid is refined using the Zhang-Suen-Thinning algorithm to obtain a skeleton line similar to the road centerline. The extraction of the skeleton line ensures accurate identification of intersection points based on neighborhood relationships. Intersection points and endpoints can be autonomously and accurately identified directly based on the 8-neighborhood relationships of the skeleton line, eliminating the dependence on intersection perception models. Furthermore, when generating a Link, it is immediately linked to the intersection point, and the directionality of the Link is determined based on the directional voting of the temporal trajectory.

[0054] Those skilled in the art will understand that, in the above-described method of the specific implementation, the order in which each step is written does not imply a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of each step should be determined by its function and possible internal logic.

[0055] Please see Figure 5 , Figure 5 This is a schematic diagram of the structure of an electronic device according to an embodiment of this application. The electronic device 50 includes a memory 51 and a processor 52 coupled to each other. The processor 52 is used to execute program instructions stored in the memory 51 to implement the steps of the road network topology map generation method embodiment described above. In a specific implementation scenario, the electronic device 50 may include, but is not limited to, a microcomputer or a server.

[0056] Specifically, processor 52 controls itself and memory 51 to implement the steps of the above-described method embodiment for generating a road network topology map. Processor 52 can also be called a CPU (Central Processing Unit), and may be an integrated circuit chip with signal processing capabilities. Processor 52 can also be a general-purpose processor, digital signal processor (DSP), application-specific integrated circuit (ASIC), field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. A general-purpose processor can be a microprocessor or any conventional processor. Furthermore, processor 52 can be implemented using integrated circuit chips.

[0057] Please see Figure 6 , Figure 6 This is a schematic diagram of the structure of a non-volatile computer-readable storage medium according to an embodiment of this application. The non-volatile computer-readable storage medium 60 is used to store a computer program 601. When the computer program 601 is executed by a processor, for example by the aforementioned... Figure 5 When the processor 52 in the embodiment is executed, it is used to implement the steps of the above-described embodiment of the method for generating road network topology maps.

[0058] The description of the various embodiments above tends to emphasize the differences between the various embodiments. The similarities or similarities between them can be referred to, and for the sake of brevity, they will not be repeated here.

[0059] In the several embodiments provided in this application, it should be understood that the disclosed methods and related devices can be implemented in other ways. For example, the related device implementations described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication disconnection shown or discussed may be indirect coupling or communication disconnection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0060] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0061] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods of various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0062] Those skilled in the art will readily recognize that numerous modifications and variations can be made to the apparatus and method while maintaining the teachings of this application. Therefore, the above disclosure should be considered limited only by the scope of the appended claims.

Claims

1. A method for generating a road network topology map, characterized in that, include: Obtain vehicle trajectory data corresponding to the parking lot, wherein the parking lot includes at least two floors; The vehicle trajectory data is processed to obtain at least two two-dimensional grid maps corresponding to the at least two floors; Identify multiple key nodes in each of the two-dimensional mesh diagrams, generate link information corresponding to the multiple key nodes, and determine the link direction information corresponding to the link information; The road network topology map is generated based on the multiple key nodes, link information, and link direction information corresponding to each of the two-dimensional grid maps.

2. The method according to claim 1, characterized in that, The vehicle trajectory data is processed to obtain at least two two-dimensional grid maps corresponding to the at least two floors, including: The vehicle trajectory data is divided according to the at least two floors to obtain sub-trajectory data corresponding to each floor; The sub-trajectory data is downsampled using a two-dimensional grid to obtain a two-dimensional grid map corresponding to the floor, wherein the number of trajectory points in each grid in the two-dimensional grid map is greater than or equal to a preset value, and each grid represents the average three-dimensional coordinates of the trajectory points within the grid.

3. The method according to claim 1, characterized in that, Identifying multiple key nodes in the two-dimensional mesh graph includes: The original trajectory region in the two-dimensional mesh image is dilated and closed to obtain the trajectory region in the two-dimensional mesh image. The trajectory region in the two-dimensional grid is refined to obtain the skeleton line grid corresponding to the road centerline; In the skeleton line mesh, the plurality of key nodes are determined.

4. The method according to claim 3, characterized in that, The multiple key nodes include first-type nodes and second-type nodes; In the skeleton line mesh, the plurality of key nodes are determined, including: Using the neighborhood relationship of the skeleton line grid, retain the first skeleton line grid point and the second skeleton line grid point in the skeleton line grid, wherein the number of neighborhood points of the first skeleton line grid point in the skeleton line grid is equal to a first preset value, and the number of neighborhood points of the second skeleton line grid point in the skeleton line grid is greater than a second preset value, wherein the second preset value is greater than the first preset value. The first skeleton line grid point is used as the first type node, and in response to the presence of a neighborhood connectivity greater than or equal to a third preset value within a preset-size window used at the second skeleton line grid point in the skeleton line grid, the second skeleton line grid point is used as the second type node. The third skeleton line grid point in the skeleton line grid is used as the second type of node, wherein the third skeleton line grid point corresponds to the ramp connection between two adjacent floors.

5. The method according to claim 3 or 4, characterized in that, Generate link information corresponding to the multiple key nodes, including: Determine the current search point, wherein the current search point is one of the multiple key nodes; In the skeleton line mesh, the skeleton line mesh points located in the extension direction of the skeleton line mesh are determined from the neighborhood points of the current search point in a recursive manner, and then used as the next search point. In the recursive loop, the angle between the current link direction corresponding to the current search point and the direction of the next search point is less than a preset angle, wherein the current link direction represents the direction between the current search point and the previous key node; In response to the next search point being another key node among the plurality of key nodes, link information between the current search point and the next search point is generated, wherein the link information includes data between the current search point and the next search point.

6. The method according to claim 1, characterized in that, The vehicle trajectory data includes trip identifiers and timestamp information; Determining the link direction information corresponding to the link information includes: The vehicle trajectory data is classified according to the trip identifier to obtain multiple vehicle trip data. For each vehicle trip data, the sequence of trajectory points corresponding to the vehicle trip data is sorted according to the timestamp information to obtain the order in which the second type of nodes among the multiple key nodes are passed. The order in which the second type of nodes among the multiple key nodes pass is used for the link direction information.

7. The method according to claim 1, characterized in that, The vehicle trajectory data includes trip identifiers and timestamp information; Determining the link direction information corresponding to the link information includes: The vehicle trajectory data is classified according to the trip identifier to obtain multiple vehicle trip data. For each vehicle trip data, the sequence of trajectory points corresponding to the vehicle trip data is sorted according to the timestamp information to obtain the order in which the second type of nodes among the multiple key nodes are passed. In response to a situation where the proportion of vehicle travel data corresponding to the order of passing through the second type of nodes among the plurality of key nodes is greater than a preset value, the order of passing through the second type of nodes among the plurality of key nodes is used for the link direction information.

8. The method according to claim 1, characterized in that, Based on the multiple key nodes, link information, and link direction information corresponding to each of the two-dimensional grid diagrams, the road network topology map is generated, further including: Connect the key nodes located at the ramp connection points among the multiple key nodes corresponding to the two-dimensional grid diagrams of adjacent floors to use in the road network topology map.

9. An electronic device, characterized in that, The method includes a memory and a processor coupled to each other, the processor being used to execute program instructions stored in the memory to implement the method for generating a road network topology map according to any one of claims 1 to 8.

10. A non-volatile computer-readable storage medium storing program instructions thereon, characterized in that, When the program instructions are executed by the processor, they implement the method for generating a road network topology map as described in any one of claims 1 to 8.

Images