Method, computer system, and medium for generating autonomous driving maps

Abstract

The present invention relates to a method, computer system, and computer storage medium for generating an autonomous driving map. According to one aspect of the present invention, the method includes the following steps: using a road test vehicle equipped with real-time dynamic measurement equipment to collect first measurement data a single time, and processing the first measurement data to generate base map information for the autonomous driving map; using a user vehicle to collect second measurement data multiple times, and processing the second measurement data to generate supplementary information for the autonomous driving map; and generating the autonomous driving map based on the base map information and the supplementary information.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving, and more particularly to a method, a computer system, and a computer storage medium for generating an autonomous driving map. Background Art

[0002] Currently, there are two main solutions for generating autonomous driving maps. One is a solution that relies on pure visual data without relying on high-precision positioning data. The other is a solution that relies on high-precision positioning data with the help of lidar. The autonomous driving map is generated by performing three-dimensional reconstruction of the point cloud data of the road collected by the lidar.

[0003] However, the safety and reliability of map generation solutions based on purely visual data require continuous improvement and verification through practical implementation, while map generation solutions that rely on LiDAR to collect point cloud data have high hardware and mapping costs. Furthermore, due to the complex road conditions in urban areas, lateral and longitudinal positioning accuracy is even more critical to improving the safety, reliability, and user experience of autonomous driving. Furthermore, in some areas, the reliability of global navigation satellite system signals may be low or even lost, necessitating solutions that rely on high-precision positioning data. However, the mapping costs of relying solely on high-precision positioning equipment and mapping vehicles are prohibitively high, making it impossible to dynamically reflect map updates. Furthermore, adding LiDAR and high-precision positioning equipment to production vehicles also incurs additional hardware costs. Summary of the Invention

[0004] In order to solve or at least alleviate one or more of the above problems, the following technical solutions are provided.

[0005] According to a first aspect of the present invention, a method for generating an autonomous driving map is provided, the method comprising the steps of: collecting first measurement data a single time using a road test vehicle equipped with real-time dynamic measurement equipment, and processing the first measurement data to generate base map information of the autonomous driving map; collecting second measurement data multiple times using a user vehicle, and processing the second measurement data to generate supplementary information of the autonomous driving map; and generating the autonomous driving map based on the base map information of the autonomous driving map and the supplementary information of the autonomous driving map.

[0006] According to a method for generating an autonomous driving map according to an embodiment of the present invention, the first measurement data and the second measurement data include positioning data and perception data, and the perception data includes one or more of the following: stop line data, lane line data, road boundary data, lane fork and merge point data, sign data, and traffic light data.

[0007] According to the method for generating an autonomous driving map according to one embodiment of the present invention or any one of the above embodiments, processing the first measurement data to generate the base map information of the autonomous driving map includes: determining the vehicle trajectory and intersection information based on the first measurement data; dividing the vehicle trajectory into an in-intersection sub-trajectory and an out-intersection sub-trajectory according to the intersection information; converting the coordinate system from the vehicle coordinate system to the world coordinate system according to the positioning data in the first measurement data, so as to project the perception data in the first measurement data to the world coordinate system to generate multiple lane line trajectories; matching the multiple lane line trajectories to the divided in-intersection sub-trajectories and out-intersection sub-trajectories to generate matched multiple lane line trajectories; and sampling and clustering each segment of the matched multiple lane line trajectories to obtain one or more of stop line information, sign information, lane line information and road boundary information to generate the base map information of the autonomous driving map.

[0008] According to the method for generating an autonomous driving map according to one embodiment of the present invention or any one of the above embodiments, processing the first measurement data to generate base map information of the autonomous driving map further includes: clustering the sub-trajectories outside the intersection into clusters using a line segment clustering method; and generating road segment information based on the clusters to serve as the base map information of the autonomous driving map.

[0009] According to the method for generating an autonomous driving map according to one embodiment of the present invention or any one of the above embodiments, sampling and clustering each segment of the multiple lane line trajectories to obtain lane line information includes: clustering the sampling points of the lane line information using a density-based clustering method to generate point clusters; determining the main direction of the point cluster using principal component analysis or endpoint method and rotating the main direction to a horizontal direction; dividing the point clusters with a length greater than a length threshold into sub-point clusters in the horizontal direction; performing least squares fitting on the point clusters and / or sub-point clusters and then sampling again to obtain matching points and discrete points, and clustering the matching points and discrete points using a density-based clustering method to generate point clusters; and determining the lane line type based on the type of points greater than a predetermined proportion in the point cluster.

[0010] According to the method for generating an autonomous driving map according to one embodiment of the present invention or any one of the above embodiments, sampling and clustering each segment of the multiple lane line trajectories to obtain stop line information includes: determining the midpoint of the stop line based on the stop line information; clustering the midpoints of the stop lines using a density-based clustering method to generate point clusters; and selecting a corresponding stop line from each point cluster, so that the midpoint of the stop line is the center of the point cluster, and the length and direction of the stop line are the average value of multiple stop lines corresponding to the center of the point cluster.

[0011] According to the method for generating an autonomous driving map according to one embodiment of the present invention or any one of the above embodiments, processing the second measurement data to generate supplementary information for the autonomous driving map includes: determining a vehicle trajectory and intersection information based on the second measurement data; dividing the vehicle trajectory into an in-intersection sub-trajectory and an out-intersection sub-trajectory according to the intersection information; converting the coordinate system from the vehicle coordinate system to the world coordinate system according to the positioning data in the second measurement data, so as to project the perception data in the second measurement data to the world coordinate system to generate multiple lane line trajectories; matching the multiple lane line trajectories to the divided in-intersection sub-trajectories and out-intersection sub-trajectories to generate matched multiple lane line trajectories; and sampling and clustering each segment of the matched multiple lane line trajectories to obtain one or more of stop line information, sign information, lane line information, road boundary information and centerline information to generate supplementary information for the autonomous driving map.

[0012] According to the method for generating an autonomous driving map according to one embodiment or any of the above embodiments of the present invention, processing the second measurement data to generate supplementary information for the autonomous driving map further includes: determining the points where the multiple lane line trajectories and the intersection edge intersect based on the multiple lane line trajectories and the intersection information; clustering the intersection points using a density-based clustering method to obtain the entrance and exit of the intersection; and determining the centerline information corresponding to the inside and outside of the intersection, respectively.

[0013] In the method for generating an autonomous driving map according to one embodiment of the present invention or any one of the above embodiments, processing the second measurement data to generate supplementary information for the autonomous driving map further includes filtering the road boundary information and centerline information based on the Hausdorff distance and the boundary distance from the centerline.

[0014] According to a second aspect of the present invention, a computer system for generating an autonomous driving map is provided, the system comprising: a memory; a processor; and a computer program stored on the memory and running on the processor, wherein the execution of the computer program causes the following steps to be performed: collecting first measurement data once using a road test vehicle equipped with real-time dynamic measurement equipment, and processing the first measurement data to generate base map information of the autonomous driving map; collecting second measurement data multiple times using a user vehicle, and processing the second measurement data to generate supplementary information of the autonomous driving map; and generating the autonomous driving map based on the base map information of the autonomous driving map and the supplementary information of the autonomous driving map.

[0015] According to a computer system for generating an autonomous driving map according to an embodiment of the present invention, the first measurement data and the second measurement data include positioning data and perception data, and the perception data includes one or more of the following: stop line data, lane line data, road boundary data, lane fork and merge point data, sign data, and traffic light data.

[0016] According to the computer system for generating an autonomous driving map according to one embodiment of the present invention or any one of the above embodiments, the operation of the computer program also causes the following steps to be performed: determining the vehicle trajectory and intersection information based on the first measurement data; dividing the vehicle trajectory into an in-intersection sub-trajectory and an out-intersection sub-trajectory according to the intersection information; converting the coordinate system from the vehicle coordinate system to the world coordinate system according to the positioning data in the first measurement data, so as to project the perception data in the first measurement data to the world coordinate system to generate multiple lane line trajectories; matching the multiple lane line trajectories to the divided in-intersection sub-trajectories and out-intersection sub-trajectories to generate matched multiple lane line trajectories; and sampling and clustering each segment of the matched multiple lane line trajectories to obtain one or more of stop line information, sign information, lane line information and road boundary information to generate the base map information of the autonomous driving map.

[0017] According to the computer system for generating an autonomous driving map according to one embodiment or any one of the above embodiments of the present invention, the execution of the computer program further causes the following steps to be performed: clustering the sub-trajectories outside the intersection into clusters using a line segment clustering method; and generating road segment information based on the clusters to serve as the base map information of the autonomous driving map.

[0018] According to the computer system for generating an autonomous driving map according to one embodiment of the present invention or any one of the above embodiments, sampling and clustering each segment of the multiple lane line trajectories to obtain lane line information includes: clustering the sampling points of the lane line information using a density-based clustering method to generate point clusters; determining the main direction of the point cluster using principal component analysis or endpoint method and rotating the main direction to a horizontal direction; dividing the point clusters with a length greater than a length threshold into sub-point clusters in the horizontal direction; performing least squares fitting on the point clusters and / or sub-point clusters and then sampling again to obtain matching points and discrete points, and clustering the matching points and discrete points using a density-based clustering method to generate point clusters; and determining the lane line type based on the type of points greater than a predetermined proportion in the point cluster.

[0019] According to the computer system for generating an autonomous driving map according to one embodiment of the present invention or any one of the above embodiments, sampling and clustering each segment of the multiple lane line trajectories to obtain stop line information includes: determining the midpoint of the stop line based on the stop line information; clustering the midpoints of the stop lines using a density-based clustering method to generate point clusters; and selecting a corresponding stop line from each point cluster, so that the midpoint of the stop line is the center of the point cluster, and the length and direction of the stop line are the average value of multiple stop lines corresponding to the center of the point cluster.

[0020] According to the computer system for generating an autonomous driving map according to one embodiment of the present invention or any one of the above embodiments, the operation of the computer program also causes the following steps to be performed: determining the vehicle trajectory and intersection information based on the second measurement data; dividing the vehicle trajectory into an in-intersection sub-trajectory and an out-intersection sub-trajectory according to the intersection information; converting the coordinate system from the vehicle coordinate system to the world coordinate system according to the positioning data in the second measurement data, so as to project the perception data in the second measurement data to the world coordinate system to generate multiple lane line trajectories; matching the multiple lane line trajectories to the divided in-intersection sub-trajectories and out-intersection sub-trajectories to generate matched multiple lane line trajectories; and sampling and clustering each segment of the matched multiple lane line trajectories to obtain one or more of stop line information, sign information, lane line information, road boundary information and centerline information to generate supplementary information for the autonomous driving map.

[0021] According to the computer system for generating an autonomous driving map according to one embodiment of the present invention or any of the above embodiments, the operation of the computer program also causes the following steps to be performed: determining the points where the multiple lane line trajectories and the intersection edge intersect based on the multiple lane line trajectories and the intersection information; clustering the intersection points using a density-based clustering method to obtain the entrance and exit of the intersection; and determining the centerline information corresponding to the inside and outside of the intersection respectively.

[0022] According to the computer system for generating an autonomous driving map according to one embodiment of the present invention or any of the above embodiments, the operation of the computer program also causes the following steps to be performed: filtering the road boundary information and centerline information based on the Hausdorff distance and the boundary distance from the centerline.

[0023] According to a third aspect of the present invention, a computer storage medium is provided, wherein the computer storage medium includes instructions, which, when run, execute the steps of the method for generating an autonomous driving map according to the first aspect of the present invention.

[0024] According to one or more embodiments of the present invention, a solution for generating autonomous driving maps utilizes high-precision measurement data collected in a single pass by a road test vehicle equipped with real-time dynamic measurement equipment to generate the base map information for the autonomous driving map. Furthermore, it utilizes navigation-level measurement data collected multiple times by user vehicles to generate supplementary information for the autonomous driving map, thereby enabling the refinement and updating of the autonomous driving map. This solution eliminates the need for large-scale high-precision map collection and production by mapping vehicles, and does not rely on high-precision positioning devices for deployment and dynamic map updates. However, it can achieve mapping results that are substantially consistent with those achieved with high-precision maps and positioning devices. This improves mapping efficiency and reduces labor and material costs while ensuring that the autonomous driving map complies with design specifications and maintains safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The above and / or other aspects and advantages of the present invention will become clearer and easier to understand through the following description of various aspects in conjunction with the accompanying drawings, in which the same or similar elements are represented by the same reference numerals. In the accompanying drawings:

[0026] Figure 1 Flowchart of a method for generating an autonomous driving map according to one or more embodiments of the present invention.

[0027] Figure 2 Flowchart of a method for generating an autonomous driving map according to one or more embodiments of the present invention.

[0028] Figure 3 A schematic block diagram of a computer system for generating an autonomous driving map according to one or more embodiments of the present invention is provided. DETAILED DESCRIPTION

[0029] The description of the following specific embodiments is merely exemplary in nature and is not intended to limit the disclosed technology or the application and use of the disclosed technology. In addition, there is no intention to be bound by any express or implied theory presented in the foregoing technical field, background technology or the following specific embodiments.

[0030] In the following detailed description of the embodiments, numerous specific details are set forth to provide a more thorough understanding of the disclosed technology. However, it will be apparent to one of ordinary skill in the art that the disclosed technology can be practiced without these specific details. In other instances, well-known features are not described in detail to avoid unnecessarily complicating the description.

[0031] Terms such as "comprising" and "including" indicate that, in addition to the units and steps directly and explicitly stated in the specification, the technical solution of the present invention does not exclude the presence of other units and steps not directly or explicitly stated. Terms such as "first" and "second" do not indicate the order of units in terms of time, space, size, etc., but are merely used to distinguish between units.

[0032] It should be noted that, in the context of the present invention, the term "real-time kinematic measurement equipment" is also referred to as RTK (Real-time kinematic) equipment, which is a measurement equipment that obtains centimeter-level precision positioning data in real time through real-time differential positioning technology.

[0033] Hereinafter, exemplary embodiments according to the present invention will be described in detail with reference to the accompanying drawings.

[0034] Figure 1 The present invention is a flowchart of a method for generating an autonomous driving map according to an embodiment of the present invention.

[0035] like Figure 1 As shown in [1], in step S101, a road test vehicle equipped with real-time dynamic measurement equipment collects first measurement data once and processes the first measurement data to generate the base map information for the autonomous driving map. In step S103, a user vehicle collects second measurement data multiple times and processes the second measurement data to generate supplementary information for the autonomous driving map. Optionally, steps S101 and S103 can be performed in parallel to improve the efficiency of data collection and map generation.

[0036] Optionally, the first measurement data and the second measurement data may include positioning data and perception data. For example, the positioning data may include, but is not limited to, vehicle position, speed, yaw angle, and yaw rate. The perception data may include, but is not limited to, stop line data, lane line data, road boundary data, lane divergence and merging point data, sign data, and traffic light data.

[0037] In one embodiment, the first measurement data may be collected once using a road test vehicle equipped with a real-time dynamic measurement device at a data collection frequency of 20 Hz, and the second measurement data may be collected multiple times using a user vehicle at a data collection frequency of 1 Hz.

[0038] In step S105, an autonomous driving map is generated based on the base map information and the supplementary information of the autonomous driving map. Optionally, the generated autonomous driving map may include lane information (e.g., lane markings, stop lines, bifurcation and convergence points), sign information (e.g., signs, traffic lights), road connection topology information, and driving experience information (e.g., center lines, guide lines, speed limits, etc.).

[0039] The following will use Figure 2 The processing of the first measurement data and the second measurement data is described in detail with reference to the steps shown. It should be noted that the processing of the first measurement data and the second measurement data is similar, so the processing of the first measurement data is described with reference to the processing of the first measurement data. Figure 2 Each step in .

[0040] Figure 2 Flowchart of a method for generating an autonomous driving map according to one or more embodiments of the present invention.

[0041] like Figure 2 As shown in FIG, in step S201, the vehicle trajectory and intersection information are determined based on the first measurement data.

[0042] Optionally, in step S201 , the vehicle trajectory may be determined based on the positioning data in the first measurement data, and the intersection information may be further determined based on the perception data in the first measurement data and the vehicle trajectory.

[0043] Optionally, in step S201, the intersection information can be determined based on the perception data and vehicle trajectory in the first measurement data in the following manner: determine the midpoint of the stop line; cluster the midpoints of the stop line using a density-based clustering method to generate point clusters; select the center of the minimum outer circle of each point cluster as the intersection center and set the minimum allowable radius of each intersection based on the radius of the minimum outer circle; detect whether there is overlap in the boundaries of the intersection, and if so, reallocate the radius in proportion to the number of midpoints of the stop lines contained in each, until it is detected that there is no overlap in the boundaries of the intersection. It should be noted that if the reallocated radius is less than the minimum allowable radius of the intersection, the radius is reallocated based on the minimum allowable radius. By selecting the center of the minimum outer circle of each point cluster as the intersection center, it is possible to prevent errors in intersection position detection caused by an imbalance in the number of stop lines detected in different branches of the same intersection.

[0044] In step S203, the vehicle trajectory is divided into sub-trajectories within the intersection and sub-trajectories outside the intersection based on the intersection information. Alternatively, the vehicle trajectory can be divided into sub-trajectories within the intersection and sub-trajectories outside the intersection based on the minimum allowable radius of each intersection. Alternatively, a line segment clustering method can be used to cluster the sub-trajectories outside the intersection into clusters, and segment information can be generated based on the clusters to serve as the base map information for the autonomous driving map.

[0045] In step S205, the coordinate system is converted from the vehicle coordinate system to the world coordinate system based on the positioning data in the first measurement data. The perception data in the first measurement data is projected into the world coordinate system to generate multiple lane track segments. For example, the perception data corresponding to each frame (e.g., lane lines, signs, etc.) can be projected into the world coordinate system to generate a lane track segment.

[0046] In step S207, the multiple lane line trajectories are matched to the divided in-intersection sub-trajectories and out-intersection sub-trajectories to generate matched multiple lane line trajectories. Optionally, an iterative closest point matching algorithm can be used to match the multiple lane line trajectories to the divided in-intersection sub-trajectories and out-intersection sub-trajectories. By matching the multiple lane line trajectories to the in-intersection sub-trajectories and out-intersection sub-trajectories divided by the vehicle trajectory, errors in the generated lane line trajectories caused by missing or unclear lane lines at the intersection can be avoided.

[0047] Optionally, in step S207, self-alignment and matching alignment can be performed on the multiple lane line segments. In self-alignment, each lane line segment in the multiple lane line segments can be aligned with the lane lines captured in the previous and next frames in the world coordinate system. In matching alignment, each lane line segment in the multiple lane line segments, the lane lines captured multiple times by the user vehicle, and the currently optimal map lane lines can be matched and aligned.

[0048] In step S209, each of the multiple lane line segments is sampled and clustered to obtain one or more of stop line information, sign information, lane line information, and road boundary information to generate base map information for the autonomous driving map. Optionally, sampling each of the multiple lane line segments may include selecting a number of points for each segment to perform curve fitting.

[0049] In one embodiment, sampling and clustering each segment of a plurality of lane line trajectories to obtain lane line information may include: clustering the lane line information sampling points using a density-based clustering method to generate point clusters; determining the main direction of the point clusters using principal component analysis or an endpoint method and rotating the main direction to be horizontal (e.g., parallel to the x-axis); horizontally segmenting point clusters whose length exceeds a length threshold into sub-clusters; performing least squares fitting on the point clusters and / or sub-clusters, then sampling again to obtain matching points and discrete points, and clustering the matching points and discrete points using a density-based clustering method to generate point clusters; and determining the lane line type based on the type of points in the point clusters that have a greater than predetermined proportion. It should be noted that the above clustering process using the density-based clustering method and the least squares fitting process on the point clusters and / or sub-clusters may be repeated to improve the accuracy of the acquired lane line information.

[0050] In one embodiment, sampling and clustering each segment of a plurality of lane line trajectories to obtain stop line information may include: determining a midpoint of the stop line based on the stop line information; clustering the midpoints of the stop lines using a density-based clustering method to generate point clusters; and selecting a corresponding stop line from each point cluster such that the midpoint of the corresponding stop line is the center of the point cluster, and the length and direction of the corresponding stop line are the average values ​​of the multiple stop lines corresponding to the center of the point cluster.

[0051] It should be noted that during the processing of the second measurement data to generate supplementary information for the autonomous driving map, each of the multiple lane line trajectories must be sampled and clustered to obtain centerline information to generate the supplementary information for the autonomous driving map. Optionally, during the processing of the second measurement data to generate supplementary information for the autonomous driving map, the obtained road boundary and centerline information can be filtered based on the Hausdorff distance and the boundary distance to the centerline to remove road boundaries and centerlines that clearly do not belong to the same road.

[0052] Optionally, in the process of sampling and clustering each segment of the multiple lane line trajectories to obtain centerline information, the points where the multiple lane line trajectories and the intersection edge intersect can be determined based on the multiple lane line trajectories and the intersection information; the intersection points can be clustered using a density-based clustering method to obtain the entrance and exit of the intersection; and the centerline information corresponding to the inside and outside of the intersection can be determined respectively.

[0053] In one embodiment, the centerline information corresponding to the intersection can be determined in the following manner: numbering the entrances and exits of the intersection; using a clustering method to group the entrances and exits of the intersection according to the road branches they are located in; in each branch, the entrances are numbered in an increasing counterclockwise direction and the exits are numbered in an increasing clockwise direction; discarding trajectories that are interrupted in the intersection and retaining trajectories that pass through the intersection; classifying the retained trajectories into clusters based on the entrances and intersections of their corresponding intersections; extracting a centerline from each cluster of trajectories using resampling and median position methods, and associating it with the entrance and intersection of the corresponding intersection.

[0054] In one embodiment, centerline information corresponding to the intersection can be determined by: using principal component analysis to find the principal direction of each trajectory point and fitting a guideline; drawing perpendicular lines at fixed distances along the guideline to divide the trajectory into subtrajectories using the perpendicular lines; clustering the intersections of the perpendicular lines and the trajectory to generate a node at the center of each cluster, thereby dividing the trajectory into subsegments connected by nodes; grouping the subsegments according to their leading node; and generating a centerline for each group of subsegments using the same method used to determine the centerline within the intersection. Optionally, the number of incoming and outgoing centerlines for each node can be analyzed to determine the node type. If a node has only one incoming centerline and one outgoing centerline, the node is deleted and its two connected centerlines are merged in sequence. If a node has more incoming centerlines than outgoing centerlines, the node is determined as a confluence point. If a node has less than or equal to the number of outgoing centerlines, the node is determined as a bifurcation point.

[0055] According to one aspect of the present invention, a method for generating autonomous driving maps does not require mapping vehicles to conduct large-scale high-precision map collection and mapping, nor does it require reliance on high-precision positioning devices during deployment and dynamic map updates. However, it can achieve a mapping effect that is basically consistent with high-precision maps and positioning devices. This improves mapping efficiency and reduces manpower and material costs while ensuring that the autonomous driving map complies with design specifications and safety.

[0056] Figure 3 A schematic block diagram of a computer system for generating an autonomous driving map according to one or more embodiments of the present invention is provided.

[0057] like Figure 3 As shown in , the computer system 30 for generating an autonomous driving map includes a communication unit 310, a memory 320 (for example, a non-volatile memory such as a flash memory, a read-only memory, a hard disk drive, a magnetic disk, or an optical disk), a processor 330, and a computer program 340 stored on the memory 320 and executable on the processor 330.

[0058] The communication unit 310 serves as a communication interface and is configured to establish a communication connection between the computer system 30 and an external device or network (eg, a mobile phone, a cloud, a remote server, etc.).

[0059] The memory 320 stores a computer program 340 executable by the processor 330. The processor 330 is configured to execute the computer program 340 to implement the method for generating an autonomous driving map according to one or more embodiments of the present invention.

[0060] In addition, as described above, the present invention may also be implemented as a computer storage medium in which a program for causing a computer to execute the method for generating an autonomous driving map according to one aspect of the present invention is stored.

[0061] Here, as computer storage media, various types of computer storage media can be used, such as disks (for example, magnetic disks, optical disks, etc.), cards (for example, memory cards, optical cards, etc.), semiconductor memories (for example, read-only memories, non-volatile memories, etc.), and tapes (for example, magnetic tapes, cassette tapes, etc.).

[0062] In the case of applicable, hardware, software or a combination of hardware and software can be used to realize the various embodiments provided by the present invention. Moreover, in the case of applicable, without departing from the scope of the present invention, the various hardware components and / or software components set forth herein can be combined into composite components comprising software, hardware and / or both. In the case of applicable, without departing from the scope of the present invention, the various hardware components and / or software components set forth herein can be divided into subcomponents comprising software, hardware or both. In addition, in the case of applicable, it is contemplated that software components can be implemented as hardware components, and vice versa.

[0063] Software according to the present invention (such as program code and / or data) can be stored on one or more computer storage media. It is also contemplated that the software identified herein can be implemented using one or more general-purpose or special-purpose computers and / or computer systems, networked and / or otherwise. Where applicable, the order of the various steps described herein can be changed, combined into composite steps, and / or divided into sub-steps to provide the features described herein.

[0064] The embodiments and examples set forth herein are provided to best illustrate embodiments according to the present invention and its specific applications, and thereby enable those skilled in the art to make and use the invention. However, those skilled in the art will appreciate that the above description and examples are provided for ease of illustration and example only. The descriptions set forth are not intended to encompass all aspects of the invention or to limit the invention to the precise forms disclosed.

Claims

1. A method for generating an autonomous driving map, characterized in that: The method comprises the following steps: collecting first measurement data once using a road test vehicle equipped with a real-time dynamic measurement device, and processing the first measurement data to generate base map information of the autonomous driving map; collecting second measurement data multiple times using the user vehicle, and processing the second measurement data to generate supplementary information for the autonomous driving map; and generating the autonomous driving map based on the base map information of the autonomous driving map and the supplementary information of the autonomous driving map; Processing the first measurement data to generate base map information of the autonomous driving map includes: determining a vehicle trajectory and intersection information based on the first measurement data; Dividing the vehicle trajectory into an intra-intersection sub-trajectory and an extra-intersection sub-trajectory according to the intersection information; Converting a coordinate system from a host vehicle coordinate system to a world coordinate system based on the positioning data in the first measurement data, so as to project the perception data in the first measurement data into the world coordinate system to generate a plurality of lane line trajectories; Matching the multiple lane line trajectories to the divided intra-intersection sub-trajectories and extra-intersection sub-trajectories to generate matched multiple lane line trajectories; and Each of the matched multiple lane line trajectories is sampled and clustered to obtain one or more of stop line information, sign information, lane line information, and road boundary information to generate base map information of the autonomous driving map.

2. The method according to claim 1, wherein the first measurement data and the second measurement data include positioning data and perception data, and the perception data includes one or more of the following: stop line data, lane line data, road boundary data, lane divergence and merging point data, sign data, and signal light data.

3. The method according to claim 1, wherein processing the first measurement data to generate base map information of the autonomous driving map further comprises: Clustering the sub-trajectories outside the intersection into clusters using a line segment clustering method; as well as Road segment information is generated based on the clusters to serve as base map information of the autonomous driving map.

4. The method according to claim 1, wherein sampling and clustering each of the multiple lane line trajectories to obtain lane line information comprises: Clustering the sampling points of the lane line information using a density-based clustering method to generate point clusters; Determine the main direction of the point cluster using principal component analysis or endpoint method and rotate the main direction into a horizontal direction; Splitting point clusters whose lengths are greater than a length threshold into sub-point clusters in the horizontal direction; performing least squares fitting on the point clusters and / or sub-point clusters and then sampling again to obtain matching points and discrete points, and clustering the matching points and discrete points using a density-based clustering method to generate point clusters; as well as The lane line type is determined based on the types of points in the point cluster that are greater than a predetermined proportion.

5. The method according to claim 1, wherein sampling and clustering each of the plurality of lane line trajectories to obtain stop line information comprises: determining a midpoint of the stop line based on the stop line information; clustering the midpoints of the stop lines using a density-based clustering method to generate point clusters; as well as A corresponding stop line is selected from each point cluster, such that the midpoint of the stop line is the center of the point cluster, and the length and direction of the stop line are the average values ​​of multiple stop lines corresponding to the center of the point cluster.

6. The method according to claim 1, wherein processing the second measurement data to generate supplementary information for the autonomous driving map comprises: determining a vehicle trajectory and intersection information based on the second measurement data; Dividing the vehicle trajectory into an intra-intersection sub-trajectory and an extra-intersection sub-trajectory according to the intersection information; converting a coordinate system from a host vehicle coordinate system to a world coordinate system based on the positioning data in the second measurement data, so as to project the perception data in the second measurement data into the world coordinate system to generate a plurality of lane line trajectories; Matching the multiple lane line trajectories to the divided intra-intersection sub-trajectories and extra-intersection sub-trajectories to generate matched multiple lane line trajectories; as well as Each of the matched multiple lane line trajectories is sampled and clustered to obtain one or more of stop line information, sign information, lane line information, road boundary information, and center line information to generate supplementary information for the autonomous driving map.

7. The method according to claim 6, wherein processing the second measurement data to generate supplementary information for the autonomous driving map further comprises: Determining, based on the multiple lane line trajectories and the intersection information, a point where the multiple lane line trajectories intersect an edge of an intersection; Clustering the intersection points using a density-based clustering method to obtain an entrance and an exit of the intersection; and The centerline information corresponding to the inside and outside of the intersection is determined respectively.

8. The method according to claim 6, wherein processing the second measurement data to generate supplementary information for the autonomous driving map further comprises: The road boundary information and centerline information are filtered based on the Hausdorff distance and the boundary distance from the centerline.

9. A computer system for generating an autonomous driving map, characterized in that: The system comprises: Memory; processor; and A computer program stored on the memory and running on the processor, the execution of the computer program causing the following steps to be performed: collecting first measurement data once using a road test vehicle equipped with a real-time dynamic measurement device, and processing the first measurement data to generate base map information of the autonomous driving map; collecting second measurement data multiple times using the user vehicle, and processing the second measurement data to generate supplementary information for the autonomous driving map; and generating the autonomous driving map based on the base map information of the autonomous driving map and the supplementary information of the autonomous driving map; The execution of the computer program further causes the following steps to be performed: determining a vehicle trajectory and intersection information based on the first measurement data; Dividing the vehicle trajectory into an intra-intersection sub-trajectory and an extra-intersection sub-trajectory according to the intersection information; Converting a coordinate system from a host vehicle coordinate system to a world coordinate system based on the positioning data in the first measurement data, so as to project the perception data in the first measurement data into the world coordinate system to generate a plurality of lane line trajectories; Matching the multiple lane line trajectories to the divided intra-intersection sub-trajectories and extra-intersection sub-trajectories to generate matched multiple lane line trajectories; and Each of the matched multiple lane line trajectories is sampled and clustered to obtain one or more of stop line information, sign information, lane line information, and road boundary information to generate base map information of the autonomous driving map.

10. The system of claim 9, wherein the first measurement data and the second measurement data include positioning data and perception data, the perception data including one or more of the following: stop line data, lane line data, road boundary data, lane divergence and merging point data, sign data, and signal light data.

11. The system of claim 9, wherein the execution of the computer program further causes the following steps to be performed: Clustering the sub-trajectories outside the intersection into clusters using a line segment clustering method; and Road segment information is generated based on the clusters to serve as base map information of the autonomous driving map.

12. The system according to claim 9, wherein sampling and clustering each of the plurality of lane line trajectories to obtain lane line information comprises: Clustering the sampling points of the lane line information using a density-based clustering method to generate point clusters; Determine the main direction of the point cluster using principal component analysis or endpoint method and rotate the main direction into a horizontal direction; Splitting point clusters whose lengths are greater than a length threshold into sub-point clusters in the horizontal direction; performing least squares fitting on the point clusters and / or sub-point clusters and then sampling again to obtain matching points and discrete points, and clustering the matching points and discrete points using a density-based clustering method to generate point clusters; as well as The lane line type is determined based on the types of points in the point cluster that are greater than a predetermined proportion.

13. The system according to claim 9, wherein sampling and clustering each of the plurality of lane line trajectories to obtain stop line information comprises: determining a midpoint of the stop line based on the stop line information; clustering the midpoints of the stop lines using a density-based clustering method to generate point clusters; as well as A corresponding stop line is selected from each point cluster, such that the midpoint of the stop line is the center of the point cluster, and the length and direction of the stop line are the average values ​​of multiple stop lines corresponding to the center of the point cluster.

14. The system of claim 9, wherein the execution of the computer program further causes the following steps to be performed: determining a vehicle trajectory and intersection information based on the second measurement data; Dividing the vehicle trajectory into an intra-intersection sub-trajectory and an extra-intersection sub-trajectory according to the intersection information; converting a coordinate system from a host vehicle coordinate system to a world coordinate system based on the positioning data in the second measurement data, so as to project the perception data in the second measurement data into the world coordinate system to generate a plurality of lane line trajectories; Matching the multiple lane line trajectories to the divided intra-intersection sub-trajectories and extra-intersection sub-trajectories to generate matched multiple lane line trajectories; as well as Each of the matched multiple lane line trajectories is sampled and clustered to obtain one or more of stop line information, sign information, lane line information, road boundary information, and center line information to generate supplementary information for the autonomous driving map.

15. The system of claim 14, wherein the execution of the computer program further causes the following steps to be performed: Determining, based on the multiple lane line trajectories and the intersection information, a point where the multiple lane line trajectories intersect an edge of an intersection; Clustering the intersection points using a density-based clustering method to obtain an entrance and an exit of the intersection; and The centerline information corresponding to the inside and outside of the intersection is determined respectively.

16. The system of claim 14, wherein the execution of the computer program further causes the following steps to be performed: The road boundary information and centerline information are filtered based on the Hausdorff distance and the boundary distance from the centerline.

17. A computer storage medium, characterized in that The computer storage medium comprises instructions that, when executed, perform the method according to any one of claims 1 to 8.

Images