Method, device, equipment and storage medium for selecting 3D coordinate points of static maps

Abstract

The present application relates to a method, apparatus, device, and storage medium for selecting 3D coordinate points in a static map. The method comprises: obtaining the positional relationship between the autonomous driving vehicle and the lane, and obtaining the height information of a first surface area in the static map based on the positional relationship; projecting a viewing angle ray onto the static map according to a preset projection direction to obtain a projection intersection, and obtaining the height information of a second surface area in the static map based on the projection intersection and the corresponding lane information; obtaining the values ​​of the three-dimensional coordinate points of the static map based on the height information of the first surface area or the height information of the second surface area, and selecting the three-dimensional coordinate points. This method can solve the problem of a small range for accurately selecting three-dimensional coordinate points in the static map.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and in particular to a method, device, equipment and storage medium for selecting 3D coordinate points in a static map. Background Art

[0002] In the field of autonomous driving technology, high-precision maps can be combined with perception systems to identify traffic lights around autonomous vehicles. They can also be combined with decision-making and planning systems to obtain global path information and the information required for lane changes. With the rapid development of autonomous driving technology, people's demand for high-precision maps is also increasing.

[0003] Static maps, as the underlying layer of high-precision maps, are generally composed of three types of vector information: lane models containing semantic information, road components, and road attributes, as well as feature layers for multi-sensor positioning. They play an important role in the quality of high-precision maps. Therefore, how to accurately select three-dimensional (3D) coordinate points near lanes in static maps is one of the important technical problems that needs to be solved. Currently, the HD Road Graph (High Definition Road Graph) method can be used to obtain the values ​​of the three-dimensional coordinate axes in static maps. However, using the HD Road Graph method generates a large amount of storage data, and the memory capacity is limited. Therefore, the selection of three-dimensional coordinate points can only be performed within a certain range near the autonomous driving vehicle, resulting in a smaller range for accurately selecting three-dimensional coordinate points in the static map. Summary of the Invention

[0004] Based on this, it is necessary to provide a method, device, equipment and storage medium for selecting three-dimensional coordinate points of static maps to address the above technical problems and improve the problem of poor performance of three-dimensional coordinate point selection.

[0005] In one aspect, a method for selecting 3D coordinate points of a static map is provided, the method comprising:

[0006] Obtaining a positional relationship between the autonomous driving vehicle and the lane, and obtaining height information of a first surface area in the static map based on the positional relationship;

[0007] Projecting the viewing angle ray onto the static map according to a preset projection direction to obtain a projection intersection point, and obtaining height information of a second surface area in the static map according to the projection intersection point and corresponding lane information;

[0008] According to the height information of the first surface area or the height information of the second surface area, the three-dimensional coordinate point values ​​of the static map are obtained, and the three-dimensional coordinate point is selected.

[0009] In one embodiment, the step of obtaining a positional relationship between the autonomous driving vehicle and the lane, and obtaining height information of a first surface area in a static map based on the positional relationship includes:

[0010] Determining whether the position of the autonomous driving vehicle overlaps with the position of the lane;

[0011] If yes, obtaining the height information of the lane to obtain the height information of the first surface area;

[0012] If not, determine whether the autonomous driving vehicle has a lane in the first surface area; if so, obtain the height information of the lane to obtain the height information of the first surface area; if not, obtain the height information of the first surface area based on a high-precision dynamic map.

[0013] In one embodiment, the step of projecting a viewing ray onto the static map according to a preset projection direction to obtain a projection intersection point, and obtaining height information of a second surface area in the static map according to the projection intersection point and corresponding lane information includes:

[0014] Determining the projection direction according to a preset projection starting point and projection center point, and projecting the viewing angle ray to a second surface area of ​​the static map according to the projection direction;

[0015] Starting from the projection starting point and increasing by a preset first search step length, a plurality of the projection intersection points are obtained;

[0016] The plurality of projected intersection points are traversed, a search area is formed with the current projected intersection point as the center according to a preset search radius, a target lane is selected from the search area, and height information of the second surface area is obtained according to the position information of the target lane.

[0017] In one embodiment, the step of selecting a target lane from the search area includes:

[0018] Traversing the plurality of lanes in the search area, and determining whether position information between the current lane and the corresponding projected intersection point satisfies a preset search condition;

[0019] If so, the lane is selected to obtain the target lane.

[0020] In one embodiment, the preset search conditions include:

[0021] The Euclidean distance between the lane and the projected intersection point is less than the search radius and the difference in height information between the lane and the projected intersection point is less than a preset height threshold, wherein a mathematical expression of the height threshold includes:

[0022] E=Step / 2+10 -9

[0023] Here, Step represents a search step, and the search step includes: the first search step.

[0024] In one embodiment, the steps after obtaining the target lane include:

[0025] Obtaining the height threshold according to a preset second search step length;

[0026] Performing an addition operation between the height information of the projected intersection point corresponding to the target lane and the height threshold to obtain a second search starting point;

[0027] Performing a subtraction operation between the height information of the projected intersection point corresponding to the target lane and the height threshold to obtain a second search end point;

[0028] Searching the lane begins from the second search starting point and increases according to the second search step size until the second search end point is reached, and determining whether the lane satisfies the search condition, wherein the search condition includes: the search point is within the lane; if so, obtaining the location information of the lane.

[0029] In one embodiment, the method further comprises:

[0030] sorting the multiple Euclidean distances between the multiple lanes and the projected intersection point, wherein the smaller the value of the Euclidean distance, the higher the priority of the corresponding lane;

[0031] The height information of the second surface area in the static map is obtained according to the priority of the lane.

[0032] On the other hand, a device for selecting 3D coordinate points of a static map is provided, the device comprising:

[0033] a first acquisition module, configured to acquire a positional relationship between the autonomous driving vehicle and the lane, and obtain height information of a first surface area in the static map based on the positional relationship;

[0034] a second acquisition module, configured to project a viewing angle ray onto the static map according to a preset projection direction to obtain a projection intersection point, and obtain height information of a second surface area in the static map according to the projection intersection point and corresponding lane information;

[0035] The coordinate selection module is used to obtain the three-dimensional coordinate point values ​​of the static map according to the height information of the first surface area or the height information of the second surface area, and select the three-dimensional coordinate point.

[0036] In another aspect, a computer device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the following steps are implemented:

[0037] Obtaining a positional relationship between the autonomous driving vehicle and the lane, and obtaining height information of a first surface area in the static map based on the positional relationship;

[0038] Projecting the viewing angle ray onto the static map according to a preset projection direction to obtain a projection intersection point, and obtaining height information of a second surface area in the static map according to the projection intersection point and corresponding lane information;

[0039] According to the height information of the first surface area or the height information of the second surface area, the three-dimensional coordinate point values ​​of the static map are obtained, and the three-dimensional coordinate point is selected.

[0040] In another aspect, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0041] Obtaining a positional relationship between the autonomous driving vehicle and the lane, and obtaining height information of a first surface area in the static map based on the positional relationship;

[0042] Projecting the viewing angle ray onto the static map according to a preset projection direction to obtain a projection intersection point, and obtaining height information of a second surface area in the static map according to the projection intersection point and corresponding lane information;

[0043] According to the height information of the first surface area or the height information of the second surface area, the three-dimensional coordinate point values ​​of the static map are obtained, and the three-dimensional coordinate point is selected.

[0044] The above-mentioned method, device, computer equipment and storage medium for selecting 3D coordinate points of a static map obtain the height information of the first surface area in the static map through the positional relationship between the autonomous driving vehicle and the lane; project the viewing angle ray to the static map according to a preset projection direction to obtain the projection intersection, and then obtain the height information of the second surface area in the static map based on the projection intersection and the corresponding lane information; obtain the three-dimensional coordinate point value of the static map based on the height information of the first surface area or the height information of the second surface area, and select the three-dimensional coordinate point, which can solve the problem of the small range of accurate selection of three-dimensional coordinate points of the static map. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 A diagram illustrating an application environment of a method for selecting 3D coordinate points of a static map in one embodiment;

[0046] Figure 2 Schematic diagram of a flow chart of a method for selecting 3D coordinate points of a static map in one embodiment;

[0047] Figure 3 A schematic diagram of a process for obtaining height information of a first surface area in one embodiment;

[0048] Figure 4 A schematic diagram of a process for obtaining height information of a second surface area in one embodiment;

[0049] Figure 5 A schematic diagram of a process for selecting a target lane in one embodiment;

[0050] Figure 6 A schematic diagram of a process for presetting search conditions in one embodiment;

[0051] Figure 7 FIG1 is a schematic diagram of a process after obtaining a target lane in one embodiment;

[0052] Figure 8 A schematic diagram of a process for obtaining height information of a second surface area in one embodiment;

[0053] Figure 9 is a structural block diagram of a 3D coordinate point selection device for a static map in another embodiment;

[0054] Figure 10 FIG. 1 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0056] For human drivers, navigation is mainly carried out through electronic maps, including route planning from the starting point to the end point, positioning and matching vehicles and roads, and POI (Point of Interest) retrieval. Among them, electronic maps can be called traditional topological maps. Human drivers can identify the main road of a road from traditional topological maps and understand which other roads the road can extend to. These are all topological information that human drivers can recognize. In general, the accuracy of traditional topological maps is at the meter level. Human drivers can actively judge the surrounding environment and combine the information of traditional topological maps to make decisions and judgments on driving information.

[0057] For autonomous vehicles, however, the accuracy and information contained in traditional topological maps are often insufficient and inaccurate. Autonomous vehicles require more specific information to make lane-changing decisions, such as the specific lane the vehicle is in and its current lane location; the presence of adjacent lanes; whether a lane change is permitted; lane type; whether lane markings are dashed or solid; whether the lane change distance is sufficient; and whether the destination can be reached after the lane change. To address this, researchers and technicians in the field of autonomous driving have proposed high-precision maps, which differ from traditional topological maps. These maps, combined with perception systems, can identify traffic lights around the autonomous vehicle and, in conjunction with decision-making and planning systems, provide global path information and the information needed for lane changes. For example, HD Road Graphs, part of HD maps, describe road details with centimeter-level accuracy, primarily including lane type, lane connectivity, traffic signs / traffic lights, crosswalks, road geometry, and other semantic information.

[0058] Furthermore, accurate self-position assessment and surrounding environmental awareness are particularly important for autonomous vehicles. Currently, sensors such as cameras, millimeter-wave radar, and lidar still have some limitations when it comes to environmental perception. This is particularly true in extreme weather conditions such as dust, rain, snow, and dense fog, where misjudgments or even failure are common. Even the fusion of multiple sensors cannot completely prevent this. High-precision maps, on the other hand, not only remain effective in extreme weather conditions but also offer a field of view that is not limited by occlusion, distance, or visual constraints. They complement these sensors at the perception level, providing autonomous vehicles with more reliable perception capabilities. Regarding perception information, high-precision maps not only provide autonomous vehicles with high-precision static information such as road networks, road topography, lanes, points of interest (POIs), buildings, and road signs, but also dynamic, real-time traffic information. By fusing these two types of information, a virtual driving environment is created, enabling vehicles to perceive, understand, and comprehend the environment, enabling them to plan routes and avoid congestion and traffic obstacles. From this point of view, high-precision maps are actually equivalent to a super-sensing container. On the one hand, they can assist existing sensors, and on the other hand, they can serve as a platform to meet the needs of lane-level planning, ultimately achieving dual enhancements in perception and decision-making. Optimally, based on the reconstruction of the 3D road environment, high-precision maps can help autonomous vehicles reduce their reliance on expensive sensors, significantly reduce system costs, and alleviate in-vehicle computing pressure. Furthermore, high-precision maps can also serve as a supplement and enhancement to existing autonomous driving sensors, strengthening the vehicle-side perception capabilities in the vehicle-road collaborative architecture, thereby improving intelligent connected applications and helping automakers, research institutions, and others conduct virtual testing of autonomous driving.

[0059] Static maps, as the underlying layer of high-precision maps, are generally composed of three types of vector information: lane models containing semantic information, road components, and road attributes, as well as feature layers for multi-sensor positioning. They play a vital role in the quality of high-precision maps. Therefore, accurately selecting 3D coordinate points near lanes in static maps is a key technical challenge that needs to be addressed.

[0060] The present application provides a method for selecting 3D coordinate points in a static map, which can be applied to Figure 1In the application environment shown. The terminal 102 communicates with the server 104 via the network. For example, the method for selecting 3D coordinate points of a static map provided in the present application can be applied to a scenario where 3D coordinate points of a static map are accurately selected. Currently, the HD Road Graph method can be used to obtain the values ​​of the three-dimensional coordinate axes in a static map, but the use of the HD RoadGraph method will generate a large amount of storage data, and the memory capacity is limited. Therefore, 3D coordinate points can only be selected within a certain range near the autonomous driving vehicle, resulting in a smaller range for accurate selection of 3D coordinate points of the static map. Therefore, this method first obtains the positional relationship between the autonomous driving vehicle and the lane, and then obtains the height information of the first surface area in the static map based on the positional relationship; then projects the viewing ray to the static map according to the preset projection direction to obtain the projection intersection, and obtains the height information of the second surface area in the static map based on the projection intersection and the corresponding lane information; finally, based on the height information of the first surface area and the height information of the second surface area, obtains the three-dimensional coordinate point values ​​of the static map, and selects the three-dimensional coordinate points, which can solve the problem of the small range of accurate selection of the three-dimensional coordinate points of the static map. For example, the environmental parameters around the autonomous driving vehicle can be obtained at the terminal 102, and the data can be processed and analyzed at the server 104, and the processed information can be transmitted to the terminal 102, wherein the terminal 102 can be but is not limited to various personal computers, laptops, smart phones, tablets, portable wearable devices or sub-servers, and the server 104 can be implemented as an independent server or a server cluster composed of multiple servers or a cloud computing platform.

[0061] In one embodiment, Figure 2 As shown, a method for selecting 3D coordinate points of a static map is provided, comprising the following steps:

[0062] S1: Obtaining a positional relationship between the autonomous driving vehicle and the lane, and obtaining height information of a first surface area in a static map based on the positional relationship;

[0063] S2: Projecting the viewing angle ray onto the static map according to a preset projection direction to obtain a projection intersection point, and obtaining height information of a second surface area in the static map according to the projection intersection point and corresponding lane information;

[0064] S3: Obtaining three-dimensional coordinate point values ​​of the static map according to the height information of the first surface area or the height information of the second surface area, and selecting the three-dimensional coordinate point.

[0065] Through the above steps, the static ground Figure 3 The problem of small selection range of D coordinate points.

[0066] To obtain the height information of the first surface area, in step S1, the height information of the first surface area in the static map is obtained based on the positional relationship between the autonomous vehicle and the lane. For example, the autonomous vehicle can be positioned using the Global Navigation Satellite System (GNSS) to obtain the position information of the autonomous vehicle. The height information of the first surface area is then obtained based on the positional relationship between the autonomous vehicle and the lane. Specifically, the first surface area is defined as: a circular area with a radius of 30 meters, centered on the autonomous vehicle; or a closed or open area of ​​any shape (including but not limited to a rectangle, an ellipse, etc.) with a maximum length of 30 meters. It is determined whether the first surface area intersects with the lane. If an intersection exists, the Z-axis value of the intersection is used as the height information of the first surface area. Furthermore, in this embodiment and other embodiments, the height information can be the Z-axis value in a Cartesian coordinate system, that is, the forward and backward travel direction of the autonomous vehicle is used as the X-axis value, the left and right travel direction is used as the Y-axis value, and the direction where the road height changes, such as an overpass, is used as the Z-axis value. In the specific implementation process, the X-axis, Y-axis, and Z-axis are merely different naming methods for different directions in the Cartesian coordinate system, and the directions of vehicles and roads can also be described in various naming methods. In other implementation processes, any coordinate system can also be selected to describe the spatial position information of vehicles and roads, which is not limited here.

[0067] In order to obtain the height information of the second surface area, in step S2, it is exemplified that the viewing angle ray is projected onto the static map according to the preset projection direction to obtain the projection intersection point, and the height information of the second surface area in the static map is obtained according to the projection intersection point and the corresponding lane information. For example, in the 3D scene of autonomous driving, it is usually necessary to select each lane (Lane). The main method of selecting objects in the three-dimensional scene usually includes: obtaining the projection intersection point by using the ray formed by the three-dimensional coordinates converted from the camera position and the line of sight dragging position. For the case where the height information is not considered, the coordinate z=0 of all height information can be assumed. When selecting a lane, the coordinate point of z=0 on the ray can be obtained as the projection intersection point, and then the nearest lane can be obtained through the intersection point to obtain the lane information. However, after the height information is introduced, the intersection cannot be obtained directly, so it can be converted into: solving the first projection intersection formed by projecting the viewing ray onto the static map; optionally, in some implementation processes, the projection height and projection angle are set according to the camera position and the line of sight drag position to determine the projection direction, and then the viewing ray is projected onto the second surface area of ​​the static map to obtain the projection intersection, and the height information (Z-axis value) of the second surface area in the static map is obtained according to the coordinate value of the first projection intersection. In some implementation processes, the definition of the second surface area can be: with the autonomous driving vehicle as the center, and then setting 100 meters as the radius to form a circular area; or, setting 100 meters as the maximum length to form a closed area or open area of ​​any shape (including but not limited to rectangle, ellipse, etc.).

[0068] In order to select the 3D coordinate points of the static map, in step S3, it is exemplified that the three-dimensional coordinate point values ​​of the static map are obtained according to the height information of the first surface area and the height information of the second surface area, and the three-dimensional coordinate points are selected. For example, after obtaining the height information of the first surface area around the autonomous driving vehicle, the 3D coordinate points of the lanes in the first surface area can be located, so as to realize the selection of the 3D coordinate points in the static map. However, when the intersection cannot be obtained in the first surface area, the 3D coordinate points of the lanes in the second surface area can be located according to the height information of the second surface area after obtaining the height information of the second surface area around the autonomous driving vehicle, so as to realize the selection of a larger range of 3D coordinate points in the static map.

[0069] In some embodiments, as Figure 3 As shown, the steps of obtaining the positional relationship between the autonomous driving vehicle and the lane and obtaining the height information of the first surface area in the static map according to the positional relationship include:

[0070] S11: Determine whether the position of the autonomous driving vehicle overlaps with the position of the lane;

[0071] S12: If yes, obtaining the height information of the lane to obtain the height information of the first surface area;

[0072] S13: If not, determine whether the autonomous driving vehicle has a lane in the first surface area; if so, obtain the height information of the lane to obtain the height information of the first surface area; if not, obtain the height information of the first surface area based on a high-precision dynamic map.

[0073] like Figure 3 As shown, in steps S11 to S13, the height information of the first surface area is obtained based on the overlap between the autonomous vehicle and the lane. For example, the position of the autonomous vehicle can be accurately located using methods such as a global navigation satellite system, an inertial measurement unit, and point cloud positioning, thereby meeting the requirements of the autonomous vehicle, such as centimeter-level accuracy and millisecond-level time delay. Among them, the global navigation satellite system generally has a meter-level accuracy, and the global navigation satellite system can be brought to the centimeter level through RTK (Real-Time Kinematic) carrier phase difference technology; the inertial measurement unit obtains acceleration and angular velocity through a three-directional accelerometer and a three-axis gyroscope, which has a certain cumulative error. The cumulative error of the inertial measurement unit can be eliminated to a certain extent by the inertial navigation system; the point cloud positioning method has good robustness and does not rely on GPS (Global Positioning System) information, but relies on a priori maps. The accuracy and real-time nature of the map may cause positioning errors. Therefore, it is preferred to fuse multiple sensors and improve the accuracy and robustness of the overall positioning effect based on the Kalman filter model. After locating the position of the autonomous driving vehicle, it can be determined whether the autonomous driving vehicle overlaps with the lane: if the autonomous driving vehicle overlaps with the lane, the height information (Z-axis value) of the overlapping lane is used as the height information of the first surface area; if the autonomous driving vehicle does not overlap with the lane, it is determined whether there is a lane near the autonomous driving vehicle. If so, the height information of the lane is used as the height information of the first surface area. If not, the height information of the first surface area is obtained through the HD Road Graph (high-precision road graph) method. During the implementation process, a large amount of data will be calculated and stored when using the HDRoad Graph method, so the HD Road Graph method is only used in a part of the range.

[0074] In order to obtain the height information of the second surface area, such as Figure 4As shown, in some embodiments, the step of projecting the viewing angle ray onto the static map according to a preset projection height and projection angle to obtain a projection intersection, and obtaining height information of the second surface area in the static map according to the projection intersection includes:

[0075] S21: determining the projection direction according to a preset projection starting point and projection center point, and projecting the viewing angle ray to a second surface area of ​​the static map according to the projection direction;

[0076] S22: Starting from the projection starting point and increasing by a preset first search step length, a plurality of projection intersection points are obtained;

[0077] S23: Traverse the multiple projection intersection points, form a search area with the current projection intersection point as the center according to a preset search radius, select a target lane from the search area, and obtain the height information of the second surface area according to the position information of the target lane.

[0078] like Figure 4As shown, in steps S21 to S23, it is exemplarily explained that the projection direction is determined according to the projection starting point and the projection center point, the viewing angle ray is projected to the second surface area according to the projection direction, and multiple projection intersections are selected according to the first search step, and search areas are formed around the projection intersections in turn, and the target lane is selected from the search area. For example, the projection starting point and the projection center point can be determined by the three-dimensional coordinate shape converted from the camera position and the line of sight dragging position, and the projection direction of the viewing angle ray is formed from the projection starting point to the outside of the projection center point. The viewing angle ray is projected to the second surface area to obtain the projection intersection, so as to select the object in the three-dimensional scene. In some implementations, since the static map data introduces height information, the height information of each lane may be different, resulting in the inability to directly determine the height information of the intersection of the viewing ray and the second surface area. Therefore, the normalized vector of the viewing ray can be obtained, and the Z-axis value corresponding to the camera center can be used to obtain the three-dimensional coordinate point. Since the position of the camera center is always lower than the camera's viewing angle, it can be ensured that the direction of the viewing ray is always offset downward from the camera's viewing angle. Then, according to the pre-set first search step, starting from the projection starting point, the first search step is incremented in sequence to obtain multiple projection intersections. In some implementations, the first search step size Step_z1 can be set to 1 meter, that is, a projection intersection point is selected every 1 meter, and then each projection intersection point is traversed, with the current projection intersection point as the center, and a search area is formed according to the preset search radius. It should be noted that during the traversal of the projection intersection points, projection intersection points are selected every 1 meter starting from the projection starting point. When a projection intersection point meets the search conditions, the traversal is stopped, and a second search step size with a smaller value is used to perform a subdivision search starting from the current projection intersection point. If the search conditions are met during the subdivision search process, further search is stopped. In some implementations, the search radius can be set to 20 meters. It should be noted that the search area can be circular, or any regular, irregular, closed, or semi-closed area shape such as elliptical or rectangular, and is not limited here. A target lane that meets the conditions is selected within the search area, and the height information of the second surface area is obtained based on the position information of the target lane.

[0079] In some embodiments, as Figure 5 As shown, the step of selecting a target lane from the search area includes:

[0080] S31: traversing the plurality of lanes in the search area to determine whether the position information between the current lane and the corresponding projected intersection point satisfies a preset search condition;

[0081] S32: If yes, select the lane to obtain the target lane.

[0082] like Figure 5 As shown, in steps S31 to S32, it is exemplarily explained that a target lane is selected from multiple lanes based on a judgment result of whether the position information of the multiple lanes and the projected intersection meets the requirements. For example, for each search area, there may be multiple lanes in the search area, but not every lane meets the requirements. Therefore, it is necessary to judge whether the position information between the multiple lanes and the projected intersection meets the preset search conditions. If so, the corresponding lane is selected as the target lane.

[0083] In some embodiments, as Figure 6 As shown,

[0084] The preset search conditions include:

[0085] S41: The Euclidean distance between the lane and the projected intersection point is less than the search radius and the difference in height information between the lane and the projected intersection point is less than a preset height threshold, wherein the mathematical expression of the height threshold includes:

[0086] E=Step / 2+10 -9

[0087] Wherein, Step represents the search step length, and the search step length includes: the first search step length. Figure 6 As shown, in step S41, it is exemplarily explained that it is determined whether the Euclidean distance between the lane and the projected intersection point is less than the search radius and whether the difference in height information between the lane and the projected intersection point is less than a preset height threshold. For example, the Euclidean distance D between the lane and the projected intersection point is calculated, and the mathematical expression can be: Here, x0 and y0 represent the values ​​of the projected intersection on the X and Y axes, respectively, while xi and yi represent the values ​​of the i-th point on a lane that meets the conditions on the X and Y axes, respectively. This method can be used to calculate the Euclidean distance between each lane and the projected intersection. Then, the height difference between each lane and the projected point is calculated to determine whether the difference is less than the preset height threshold E. The mathematical expression of E can be: E = Step / 2 + 10 -9 , when the search step length Step is the first search step length of 1 meter, the height threshold E is equal to (0.5+10 -9 ) meters; and the method for calculating the height information difference between the lane and the projection point is to subtract the value of the lane on the Z-axis coordinate from the value of the projection intersection point on the Z-axis coordinate and calculate the absolute value, and then compare the absolute value with the height threshold. If the absolute value is less than the height threshold and the Euclidean distance is less than the current search radius, it is considered that the lane meets the search conditions and can be used as the target lane.

[0088] In some embodiments, as Figure 7As shown, the steps after obtaining the target lane include:

[0089] S51: Obtaining the height threshold according to a preset second search step length;

[0090] S52: performing an addition operation between the height information of the projected intersection point corresponding to the target lane and the height threshold to obtain a second search starting point;

[0091] S53: performing a subtraction operation between the height information of the projected intersection point corresponding to the target lane and the height threshold to obtain a second search end point;

[0092] S54: Searching the lane from the second search starting point, increasing the search step size according to the second search step size until the second search end point is reached, and determining whether the lane satisfies the search conditions, wherein the search conditions include: the search point is within the lane; if so, obtaining the location information of the lane.

[0093] like Figure 7 As shown, in step S51 to step S54, it is exemplarily explained that according to the preset second search step, the height threshold corresponding to the second search step is obtained, and then the second search starting point and the second search end point are obtained according to the height threshold, and the search is performed between the second search starting point and the second search end point according to the second search step. For example, the second search step can be set to 0.01 meters, and the height threshold E corresponding to the second search step is (0.005+10 -9 ) meters. Let's assume the height of the projected intersection point is Z. Then, the value of the second search starting point is Z+E, and the value of the second search end point is ZE. During the search, the projected intersection point between the second search starting point and the second search end point is obtained according to the second search step size. A determination is made as to whether the projected intersection point is within the lane. Specifically, the distance from the projected intersection point to the lane centerline is determined to be less than half the lane width. The search stops when the first lane meeting the conditions is found. This approach allows for a rough search to be performed initially with a first search step size of 1 meter. If a lane meets the conditions, a fine search and precise positioning can be performed with a second search step size of 0.01 meters. This ensures accuracy while reducing computational complexity.

[0094] In some embodiments, as Figure 8 As shown, the method further includes:

[0095] S61: sorting the multiple Euclidean distances between the multiple lanes and the projected intersection point, wherein the smaller the value of the Euclidean distance, the higher the priority of the corresponding lane;

[0096] S62: Obtaining height information of a second surface area in the static map according to the priority of the lane.

[0097] like Figure 8 As shown, in steps S61 to S62, it is exemplarily explained that multiple Euclidean distances between multiple lanes and the projected intersection are sorted, and the smaller the value of the Euclidean distance, the higher the priority of the corresponding lane. The height information of the second surface area in the static map is obtained according to the priority of the lane. For example, for lanes that meet the search conditions in the search area, the Euclidean distances between each lane and the projected intersection are counted, and the Euclidean distances are sorted. The smaller the Euclidean distance value, the closer the distance between the lane and the projected intersection. The priority of the lane corresponding to the Euclidean distance value is set higher, and the position information of the lane with higher priority can be obtained preferentially subsequently, so as to obtain the height information of the second surface area in the static map.

[0098] In some implementation processes, different search step sizes are selected multiple times to obtain multiple projection intersection points of the view ray on multiple lanes in the static map. Based on the multiple projection intersection points, the position information of the multiple lanes can be obtained. Preferably, for scenes such as overpasses, there may be two overlapping lanes on the Z axis. If the original method that does not consider height information is used, a lane will be randomly selected from the overlapping lanes. After adopting this method, lanes with different height information can be accurately selected based on the height information, that is, the lanes overlapping on the Z axis have different height information and the same horizontal information, where the horizontal information can be the values ​​of the X axis and the Y axis. Through this method, the 3D coordinate point values ​​of the static map can be obtained.

[0099] In one embodiment, Figure 9 As shown, a 3D coordinate point selection device for a static map is provided, which is characterized by comprising:

[0100] a first acquisition module, configured to acquire a positional relationship between the autonomous driving vehicle and the lane, and obtain height information of a first surface area in the static map based on the positional relationship;

[0101] a second acquisition module, configured to project a viewing angle ray onto the static map according to a preset projection direction to obtain a projection intersection point, and obtain height information of a second surface area in the static map according to the projection intersection point and corresponding lane information;

[0102] The coordinate selection module is used to obtain the three-dimensional coordinate point values ​​of the static map according to the height information of the first surface area or the height information of the second surface area, and select the three-dimensional coordinate point.

[0103] In the first acquisition module, it is exemplified that the height information of the first surface area in the static map is obtained based on the positional relationship between the autonomous driving vehicle and the lane. For example, the autonomous driving vehicle can be positioned through a global navigation satellite system or a point cloud system to obtain the position information of the autonomous driving vehicle. Then, the height information of the first surface area is obtained based on the positional relationship between the position information of the autonomous driving vehicle and the lane. Specifically, the definition of the first surface area includes: taking the autonomous driving vehicle as the center and setting a radius of 30 meters to form a circular area; or setting a maximum length of 30 meters to form a closed area or open area of ​​any shape (including but not limited to a rectangle, ellipse, etc.). It is determined whether there is an intersection between the first surface area and the lane. If there is an intersection, the Z-axis value of the intersection is used as the height information of the first surface area.

[0104] In the second acquisition module, it is exemplified that the viewing ray is projected onto the static map according to a preset projection direction to obtain a projection intersection, and the height information of the second surface area in the static map is obtained according to the projection intersection and the corresponding lane information. For example, in a 3D scene of autonomous driving, it is usually necessary to select each lane, and the main method of selecting objects in the three-dimensional scene usually includes: obtaining the projection intersection by forming a ray through three-dimensional coordinates converted from the camera position and the line of sight dragging position, that is, solving the first projection intersection formed by projecting the viewing ray onto the static map; optionally, in some implementation processes, the projection starting point and the projection center point are set according to the camera position and the line of sight dragging position to determine the projection direction, and then the viewing ray is projected onto the second surface area of ​​the static map according to the projection direction to obtain the projection intersection, and the height information of the second surface area in the static map is obtained according to the coordinate value of the first projection intersection.

[0105] In the coordinate selection module, it is exemplified that the three-dimensional coordinate point values ​​of the static map are obtained according to the height information of the first surface area or the height information of the second surface area, and the three-dimensional coordinate points are selected. For example, after obtaining the height information of the first surface area around the autonomous driving vehicle, the 3D coordinate points of the lanes in the first surface area can be located, so as to realize the selection of the 3D coordinate points in the static map. However, when the intersection point cannot be obtained in the first surface area, the 3D coordinate points of the lanes in the second surface area can be located according to the height information of the second surface area after obtaining the height information of the second surface area around the autonomous driving vehicle, so as to realize the selection of a larger range of 3D coordinate points in the static map.

[0106] The above device can be used in scenarios where 3D coordinate points of a static map are precisely selected. The first acquisition module acquires the positional relationship between the autonomous driving vehicle and the lane, and obtains the height information of the first surface area in the static map based on the positional relationship. The second acquisition module projects the viewing angle ray onto the static map according to a preset projection direction to obtain the projection intersection, and obtains the height information of the second surface area in the static map based on the projection intersection and the corresponding lane information. The coordinate selection module selects the height information of the first surface area or the height information of the second surface area to obtain the three-dimensional coordinate point values ​​of the static map, and selects the three-dimensional coordinate point. This can solve the problem of the small range for precisely selecting three-dimensional coordinate points of the static map.

[0107] The specific definition of the 3D coordinate point selection device for a static map can be found in the definition of the 3D coordinate point selection method for a static map above, and will not be repeated here. The various modules in the above-mentioned 3D coordinate point selection device for a static map can be implemented in whole or in part by software, hardware, or a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in hardware form, or can be stored in the memory of the computer device in software form, so that the processor can call and execute the corresponding operations of the above-mentioned modules.

[0108] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 10 As shown. The computer device includes a processor, a memory, a network interface and a database connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store 3D coordinate data of a static map. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a method for selecting 3D coordinate points of a static map is implemented.

[0109] Those skilled in the art will understand that Figure 10 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0110] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the following steps are performed:

[0111] Obtaining a positional relationship between the autonomous driving vehicle and the lane, and obtaining height information of a first surface area in the static map based on the positional relationship;

[0112] Projecting the viewing angle ray onto the static map according to a preset projection direction to obtain a projection intersection point, and obtaining height information of a second surface area in the static map according to the projection intersection point and corresponding lane information;

[0113] According to the height information of the first surface area or the height information of the second surface area, the three-dimensional coordinate point values ​​of the static map are obtained, and the three-dimensional coordinate point is selected.

[0114] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0115] Obtaining a positional relationship between the autonomous driving vehicle and the lane, and obtaining height information of a first surface area in the static map based on the positional relationship;

[0116] Projecting the viewing angle ray onto the static map according to a preset projection direction to obtain a projection intersection point, and obtaining height information of a second surface area in the static map according to the projection intersection point and corresponding lane information;

[0117] According to the height information of the first surface area or the height information of the second surface area, the three-dimensional coordinate point values ​​of the static map are obtained, and the three-dimensional coordinate point is selected.

[0118] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0119] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0120] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art could make various modifications and improvements without departing from the spirit of the present application, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present patent application shall be determined by the appended claims.

Claims

1. A method for selecting 3D coordinate points of a static map, characterized in that: include: Obtaining a positional relationship between the autonomous driving vehicle and the lane, and obtaining height information of a first surface area in the static map based on the positional relationship; Projecting a viewing angle ray onto the static map according to a preset projection direction to obtain a projection intersection point, and obtaining height information of a second surface area in the static map according to the projection intersection point and corresponding lane information, wherein the second surface area is larger than the first surface area; According to the height information of the first surface area or the height information of the second surface area, the three-dimensional coordinate point values ​​of the static map are obtained, and the three-dimensional coordinate point is selected.

2. The method for selecting 3D coordinate points of a static map according to claim 1, wherein: The steps of obtaining a positional relationship between the autonomous driving vehicle and the lane and obtaining height information of a first surface area in a static map based on the positional relationship include: Determining whether the position of the autonomous driving vehicle overlaps with the position of the lane; If yes, obtaining the height information of the lane to obtain the height information of the first surface area; If not, determine whether the autonomous driving vehicle has a lane in the first surface area; if so, obtain the height information of the lane to obtain the height information of the first surface area; if not, obtain the height information of the first surface area based on a high-precision dynamic map.

3. The method for selecting 3D coordinate points of a static map according to claim 1 or 2, wherein: The steps of projecting a viewing angle ray onto the static map according to a preset projection direction to obtain a projection intersection point, and obtaining height information of a second surface area in the static map according to the projection intersection point and corresponding lane information include: Determining the projection direction according to a preset projection starting point and projection center point, and projecting the viewing angle ray to a second surface area of ​​the static map according to the projection direction; Starting from the projection starting point and increasing by a preset first search step length, a plurality of the projection intersection points are obtained; The plurality of projected intersection points are traversed, a search area is formed with the current projected intersection point as the center according to a preset search radius, a target lane is selected from the search area, and height information of the second surface area is obtained according to the position information of the target lane.

4. The method for selecting 3D coordinate points of a static map according to claim 3, wherein: The step of selecting a target lane from the search area comprises: Traversing the plurality of lanes in the search area, and determining whether position information between the current lane and the currently traversed projected intersection point satisfies a preset search condition; If so, the lane is selected to obtain the target lane.

5. The method for selecting 3D coordinate points of a static map according to claim 4, wherein: The preset search conditions include: The Euclidean distance between the lane and the projected intersection point is less than the search radius and the difference in height information between the lane and the projected intersection point is less than a preset height threshold, wherein a mathematical expression of the height threshold includes: E=Step / 2+10 -9 Here, Step represents a search step, and the search step includes: the first search step.

6. The method for selecting 3D coordinate points of a static map according to claim 5, wherein: The steps after obtaining the target lane include: Obtaining the height threshold according to a preset second search step length; Performing an addition operation between the height information of the projected intersection point corresponding to the target lane and the height threshold to obtain a second search starting point; Performing a subtraction operation between the height information of the projected intersection point corresponding to the target lane and the height threshold to obtain a second search end point; Searching the lane begins from the second search starting point and increases according to the second search step size until the second search end point is reached, and determining whether the lane satisfies the search condition, wherein the search condition includes: the search point is within the lane; if so, obtaining the location information of the lane.

7. The method for selecting 3D coordinate points of a static map according to claim 1, wherein: Also includes: sorting the multiple Euclidean distances between the multiple lanes and the projected intersection point, wherein the smaller the value of the Euclidean distance, the higher the priority of the corresponding lane; The height information of the second surface area in the static map is obtained according to the priority of the lane.

8. A 3D coordinate point selection device for a static map, characterized in that: include: a first acquisition module, configured to acquire a positional relationship between the autonomous driving vehicle and the lane, and obtain height information of a first surface area in the static map based on the positional relationship; a second acquisition module, configured to project a viewing angle ray onto the static map according to a preset projection direction to obtain a projection intersection point, and obtain height information of a second surface area in the static map according to the projection intersection point and corresponding lane information, wherein the second surface area is larger than the first surface area; The coordinate selection module is used to obtain the three-dimensional coordinate point values ​​of the static map according to the height information of the first surface area or the height information of the second surface area, and select the three-dimensional coordinate point.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the method for selecting 3D coordinate points of a static map according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for selecting 3D coordinate points of a static map according to any one of claims 1 to 7 are implemented.

Images