A method, device and medium for generating an automatic driving map

Abstract

The application discloses a kind of generation method, equipment and medium of automatic driving map, method includes: determining the target area of the automatic driving map to be generated, the area image of the target area under different frames is collected by the camera sensor arranged on positioning robot;By pre-trained image recognition model, the road edge curve in the area image is identified, and the starting point and key point in the road edge curve are determined to obtain the road edge point cloud data composed of starting point and key point;The starting point of different road edge curves is connected to generate the initial road edge corresponding to the target area, and whether there is breakpoint in the initial road edge is determined;If there is, then add several smoothing points to breakpoint, according to several smoothing points, the initial road edge is smoothed to obtain the road edge of the target area after processing.

Description

Technical Field

[0001] This application relates to the field of autonomous driving technology, specifically to a method, device, and medium for generating autonomous driving maps. Background Technology

[0002] Intelligent driving technology typically requires high-precision maps to plan routes for vehicles and provide approximate locations of traffic lights at intersections. However, creating high-precision maps is a labor-intensive task. Currently, it usually involves robots scanning the area to be mapped, obtaining a point cloud map of the area, and then manually drawing continuous straight lines using vector mapping tools to mark road edges, lane lines, traffic lights, etc. This method is time-consuming and costly, and because different cartographers have varying levels of detail in their mapping, it is prone to losing road details or increasing the computational burden on autonomous driving algorithms due to excessive detail. Summary of the Invention

[0003] To address the aforementioned problems, this application proposes a method for generating autonomous driving maps, comprising:

[0004] The target area for generating the autonomous driving map is determined. Using camera sensors mounted on a positioning robot, regional images of the target area are acquired at different frames. A pre-trained image recognition model is used to identify road edge curves in the regional images, and the starting points and key points of these curves are determined to obtain road edge point cloud data composed of the starting points and key points. The starting points of different road edge curves are connected to generate the initial road edge corresponding to the target area, and it is determined whether there are any breaks in the initial road edge. If so, several smoothing points are added to the breaks, and the initial road edge is smoothed based on these smoothing points to obtain the processed road edge of the target area.

[0005] In one example, after identifying the road edge curves in the region image, the method further includes: for each frame of the region image, determining the travel direction corresponding to the positioning robot when acquiring the region image; if there are multiple road edge curves in the region image, determining the tangent lines corresponding to the multiple road edge curves and the deviation between the tangent lines and the travel direction; if the deviation is greater than a preset value, removing the corresponding road edge curve from the multiple road edge curves.

[0006] In one example, determining the starting point and key points in the road edge curve to obtain road edge point cloud data composed of the starting point and the key points specifically includes: for the road edge curve corresponding to each frame of regional image, taking the point at a preset distance from the starting point of the road edge curve along the traveling direction of the positioning robot as the key point in the road edge curve; and writing the starting point and the key points sequentially into the point cloud file according to the traveling direction and the time sequence of the regional images corresponding to the road edge curve to obtain road edge point cloud data composed of the starting points and key points of different road edge curves.

[0007] In one example, before designating a point at a preset distance from the starting point of the road edge curve along the direction of travel of the positioning robot as a key point in the road edge curve, the method further includes: obtaining the length of the road edge curve corresponding to the current frame region image; if the length of the road edge curve is lower than the preset distance, then the points on the road edge curve corresponding to the current frame region image are not written into the point cloud data.

[0008] In one example, adding several smoothing points at the breakpoint specifically includes: determining the first breakpoint and the last breakpoint of the initial road edge at the breakpoint according to the travel direction of the positioning robot; the first breakpoint is the last point on the initial road edge before the breakpoint along the travel direction, and the last breakpoint is the first point on the initial road edge after the breakpoint along the travel direction; determining two first neighboring points on the initial road edge adjacent to the first breakpoint and two second neighboring points adjacent to the last breakpoint, respectively; determining a first tangent line and a second tangent line at the breakpoint according to the first neighboring points and the second neighboring points; determining the intersection point between the first tangent line and the second tangent line, connecting the intersection point to the first breakpoint and the last breakpoint respectively, and determining a first midpoint of the connection segment between the intersection point and the first breakpoint, and a second midpoint of the connection segment between the intersection point and the last breakpoint; adding the first midpoint and the second midpoint as smoothing points to the breakpoint.

[0009] In one example, smoothing the initial road edge based on the several smoothing points to obtain the processed road edge of the target area specifically includes: generating a first position sequence consisting of the first breakpoint, the smoothing point, and the last breakpoint arranged sequentially according to the travel direction of the positioning robot; determining the third midpoint of each connecting segment formed by connecting adjacent points in the first position sequence; generating a corresponding second position sequence based on the first breakpoint, each of the third midpoints, and the first breakpoint; generating a smooth road edge corresponding to the breakpoint based on the second position sequence, and determining the inflection point in the smooth road edge; determining the curvature corresponding to the inflection point; if the curvature is greater than a preset curvature, determining the fourth midpoint of each connecting segment formed by connecting adjacent points in the second position sequence; and sequentially connecting the first breakpoint, each of the fourth midpoints, and the last breakpoint to obtain the road edge of the target area.

[0010] In one example, after obtaining the processed road edge of the target area, the method further includes: acquiring a point cloud map corresponding to the target area; determining the actual road edge corresponding to the target area based on the point cloud map; the point cloud map and the autonomous driving map are in the same coordinate system; comparing the road edge of the target area with the actual road edge to determine whether the road edge deviates from the actual road edge; if so, correcting the road edge.

[0011] In one example, after obtaining the processed road edge of the target area, the method further includes: determining the target area where the positioning robot is located as a designated area, wherein the designated area is an area with straight lane lines and right-angle curves; determining the initial direction of the road edge based on the initial distance segment of the road edge; determining the projected distance between each edge point and the initial starting point on the road edge in the initial direction by traversing each edge point on the road edge; wherein the initial starting point is the starting point of the road edge in the initial distance segment; if the projected distance corresponding to the current edge point is less than the projected distance corresponding to the previous edge point, then the current edge point is determined as a turning point, and the turning point is performed with the current edge point as the new initial starting point to obtain the straight road edge of the target area.

[0012] This application also provides an autonomous driving map generation device, characterized in that it includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to: determine a target area for generating an autonomous driving map; acquire regional images of the target area in different frames using a camera sensor mounted on a positioning robot; identify road edge curves in the regional images using a pre-trained image recognition model, and determine the starting points and key points in the road edge curves to obtain road edge point cloud data composed of the starting points and the key points; connect the starting points of different road edge curves to generate an initial road edge corresponding to the target area, and determine whether there are any breaks in the initial road edge; if so, add several smoothing points to the breaks, and smooth the initial road edge according to the several smoothing points to obtain the processed road edge of the target area.

[0013] This application also provides a non-volatile computer storage medium storing computer-executable instructions, which are configured to: determine a target area for generating an autonomous driving map; acquire regional images of the target area in different frames using a camera sensor mounted on a positioning robot; identify road edge curves in the regional images using a pre-trained image recognition model, and determine the starting points and key points of the road edge curves to obtain road edge point cloud data composed of the starting points and key points; connect the starting points of different road edge curves to generate an initial road edge corresponding to the target area, and determine whether there are any breaks in the initial road edge; if so, add several smoothing points to the breaks, and smooth the initial road edge according to the several smoothing points to obtain the processed road edge of the target area.

[0014] The method for generating autonomous driving maps proposed in this application can bring the following beneficial effects:

[0015] The system identifies road edge images within a region image and then generates initial road edges for the target region using key points in these images. This process eliminates the need for manual drawing, reducing time and labor costs and effectively improving mapping efficiency. For any breaks in the initial road edges, smoothing points can be added to these breaks, resulting in smoothed road edges. This allows for timely correction of errors that occur during mapping, effectively improving map accuracy. Attached Figure Description

[0016] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0017] Figure 1 A flowchart illustrating a method for generating an autonomous driving map provided in an embodiment of this application;

[0018] Figure 2 A schematic diagram of road edge point cloud data provided in an embodiment of this application;

[0019] Figure 3 A schematic diagram illustrating a smoothing method provided in an embodiment of this application;

[0020] Figure 4 A road edge comparison image provided for an embodiment of this application;

[0021] Figure 5 A comparison image of road edge sharpening processing provided for an embodiment of this application;

[0022] Figure 6 This is a schematic diagram of the structure of an autonomous driving map generation device provided in an embodiment of this application. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] The technical solutions provided by the various embodiments of this application are described in detail below with reference to the accompanying drawings.

[0025] like Figure 1 As shown in the embodiment of this application, a method for generating an autonomous driving map includes:

[0026] S101: Determine the target area for generating the autonomous driving map, and collect regional images of the target area in different frames using the camera sensors installed on the positioning robot.

[0027] In this embodiment, a target area for generating an autonomous driving map is determined. Before generating the autonomous driving map for the target area, the localization robot first scans the target area and generates a corresponding point cloud map. Each point in the point cloud map can locate the target area, and the coordinate system of the point cloud map is the same as that of the autonomous driving map to be generated. Thus, comparing the generated autonomous driving map with the point cloud map makes it easier to adjust and correct the corresponding autonomous driving map.

[0028] Because different roads have different conditions—for example, some roads have raised curbs—radar can be used to acquire point cloud data of the target area to improve map accuracy. Conversely, some roads have colored markings along their edges; to make these markings more prominent, a camera sensor can be used to capture regional images of the target area. The specific acquisition method can be selected according to the actual situation; the above are only some examples and are not limited in this application. In this embodiment, a camera sensor is used as the acquisition device, and regional images of the target area are acquired in different frames using a camera sensor mounted on a positioning robot.

[0029] S102: Using a pre-trained image recognition model, identify road edge curves in the regional image and determine the starting point and key points in the road edge curves to obtain road edge point cloud data composed of the starting point and key points.

[0030] After acquiring the regional image, a pre-trained image recognition model can be used to identify road edge curves within the image. Since multiple road edge curves may exist, further filtering is required to reduce computational burden. First, for each frame of the regional image, the robot's direction of travel during image acquisition must be determined. Then, the tangent lines corresponding to each of the multiple road edge curves and their deviation from the direction of travel are determined. If the deviation exceeds a preset value, the corresponding road edge curve is removed from the pool of multiple curves. Finally, one remaining road edge curve is selected for further processing.

[0031] After obtaining the road edge curve, it is necessary to determine the starting point and key points in the road edge curve in order to obtain the road edge point cloud data composed of the starting point and key points.

[0032] Specifically, for the road edge curve corresponding to each frame of the regional image, points at a preset distance from the starting point of the road edge curve along the direction of travel of the positioning robot are designated as key points in the road edge curve. Then, according to the direction of travel and the time sequence of the regional images corresponding to the road edge curve, the starting point and key points are sequentially written into the point cloud file to obtain road edge point cloud data composed of the starting points and key points of different road edge curves. In other words, the road edge point cloud data is composed of multiple starting points and key points arranged in time sequence, and the order is start-key-start-key... Figure 2 The diagram shows a road edge point cloud data. Road edge point cloud data can roughly reflect the road edge of the target area. However, due to factors such as occlusion or potholes on the road, some data may be missing or the edges may be messy. In this case, further processing of the road edge point cloud data is required to obtain a smoother road edge.

[0033] In one embodiment, before designating points at a preset distance from the starting point of the road edge curve along the robot's travel direction as key points in the road edge curve, the length of the road edge curve corresponding to the current frame region image is obtained. If the length of the road edge curve is less than the preset distance, the points on the road edge curve corresponding to the current frame region image are not written into the point cloud data. This controls the number of line segments and improves computational efficiency.

[0034] S103: Connect the starting points of different road edge curves, generate the initial road edge corresponding to the target area, and determine whether there are any breakpoints in the initial road edge.

[0035] After filtering out line segments of inconsistent length, the starting points of different road edge curves can be connected to generate the initial road edges corresponding to the target area. Since road edges may be obscured during the data acquisition process, and the obscured sections lack point clouds, the initial road edges may have partial edge gaps. Therefore, it is necessary to determine whether there are any breakpoints in the initial road edges. For example, ... Figure 3 As shown, there is a breakpoint between keypoints A and B, which causes the initial road edge to break. To fit a complete road edge, corresponding smoothing points need to be added at the breakpoint to restore the breakpoint to a smooth edge state.

[0036] S104: If it exists, add several smoothing points at the breakpoint, and smooth the initial road edge according to the several smoothing points to obtain the processed road edge of the target area.

[0037] First, based on the robot's direction of travel, determine the initial breakpoint and the final breakpoint corresponding to the breakpoint on the initial road edge. The initial breakpoint is the final point on the initial road edge before the breakpoint along the direction of travel, and the final breakpoint is the initial point on the initial road edge after the breakpoint along the direction of travel.

[0038] Then, two first neighboring points adjacent to the first end breakpoint and two second neighboring points adjacent to the last end breakpoint are determined on the initial road edge. Based on the first neighboring points and the second neighboring points, the first tangent line and the second tangent line at the breakpoint are determined accordingly.

[0039] Next, the intersection point between the first and second tangent lines is determined. This intersection point is then connected to the first and last breakpoints, respectively. The first midpoint of the connecting segment between the intersection point and the first breakpoint, and the second midpoint of the connecting segment between the intersection point and the last breakpoint are also determined. After obtaining the first and second midpoints, they are added as smoothing points to the breakpoints.

[0040] like Figure 3 As shown, the intersection point is connected to the first breakpoint A and the last breakpoint B, respectively. The first midpoint C of the connecting segment between the intersection point and the first breakpoint, and the second midpoint D of the connecting segment between the intersection point and the last breakpoint are determined. The first midpoint C and the second midpoint D are the two smoothing points added to the breakpoint. After adding the smoothing points to the breakpoint, each point (key point and smoothing point) at the breakpoint needs to be smoothed to obtain a smoother road edge.

[0041] Specifically, following the robot's direction of travel, a first point sequence [A, C, D, B] is generated, consisting of a first breakpoint, a smoothing point, and an end breakpoint arranged sequentially. The third midpoints E, F, and G of the connecting segments formed by adjacent points in the first point sequence are determined. Based on the first breakpoint, the third midpoints, and the first breakpoint, a corresponding second point sequence [A, E, F, G, B] is generated. After obtaining the second point sequence, the smooth road edge corresponding to the breakpoint is generated, and the inflection point F in the smooth road edge is determined. Then, the curvature corresponding to this inflection point is determined. If the curvature is greater than a preset curvature, it indicates that although the currently fitted road edge is smooth, its convexity is high, and further smoothing is needed to soften it. Therefore, the fourth midpoints H, I, J, and K of the connecting segments formed by adjacent points in the second point sequence are determined. By sequentially connecting the first breakpoint A, the fourth midpoints H, I, J, K, and the end breakpoint B, the smoothed road edge of the target area can be obtained.

[0042] Thus, several smoothing points are added to the breakpoints in the initial road edge to restore its integrity. By smoothing the connecting segments at the breakpoints using these smoothing points, a smooth road edge can be generated. The final road edge obtained in this embodiment is a smooth curve composed of multiple line segments. However, to ensure it reflects the actual road conditions of the target area, a point cloud map corresponding to the target area needs to be obtained. The actual road edge corresponding to the target area is determined based on the point cloud map, and then compared with the actual road edge in the point cloud map to determine whether the road edge deviates from the actual road edge. Figure 4 As shown in the figure, the circled part is where the road edge deviates from the actual road edge. In this case, the road edge needs to be corrected so that the final road edge can reflect the real road conditions and improve the map accuracy.

[0043] In one embodiment, after obtaining the processed road edge of the target area, if the target area where the positioning robot is located is a designated area (where the lane lines are straight and the curves are 90° right angles), the smoothed road edge needs to be sharpened. First, the initial direction of the road edge is determined based on the initial distance segment of the road edge. Then, by traversing each edge point on the road edge, the projected distance of each edge point in the initial direction to the initial starting point on the road edge is determined. Here, the initial starting point is the starting point of the road edge in the initial distance segment. If the projected distance corresponding to the current edge point is less than the projected distance corresponding to the previous edge point, the current edge point is determined as a turning point, and the turning point is used as the new initial starting point to obtain the straight road edge of the target area. The sharpening effect is as follows: Figure 5 As shown.

[0044] After the map is corrected and confirmed to be error-free, the road edges can be exported and saved as a CSV file using the AutowareMapToolbox tool. By viewing this file, you can obtain a high-precision autonomous driving map of the target area.

[0045] The above are embodiments of the methods proposed in this application. Based on the same idea, some embodiments of this application also provide devices corresponding to the above methods.

[0046] Figure 6 This is a schematic diagram of the structure of a data version management device based on a data warehouse, provided as an embodiment of this application. Figure 6 As shown, it includes:

[0047] At least one processor; and,

[0048] A memory that is communicatively connected to at least one processor; wherein,

[0049] The memory stores instructions that can be executed by at least one processor, and the instructions, when executed by at least one processor, enable at least one processor to:

[0050] The target area for generating the autonomous driving map is determined. Using camera sensors mounted on a positioning robot, regional images of the target area are acquired at different frames. A pre-trained image recognition model is used to identify road edge curves in the regional images, and the starting points and key points of these curves are determined to obtain road edge point cloud data composed of the starting points and key points. The starting points of different road edge curves are connected to generate the initial road edge corresponding to the target area, and it is determined whether there are any breaks in the initial road edge. If so, several smoothing points are added to the breaks, and the initial road edge is smoothed based on these smoothing points to obtain the processed road edge of the target area.

[0051] This application provides a non-volatile computer storage medium storing computer-executable instructions, which are configured as follows:

[0052] The target area for generating the autonomous driving map is determined. Using camera sensors mounted on a positioning robot, regional images of the target area are acquired at different frames. A pre-trained image recognition model is used to identify road edge curves in the regional images, and the starting points and key points of these curves are determined to obtain road edge point cloud data composed of the starting points and key points. The starting points of different road edge curves are connected to generate the initial road edge corresponding to the target area, and it is determined whether there are any breaks in the initial road edge. If so, several smoothing points are added to the breaks, and the initial road edge is smoothed based on these smoothing points to obtain the processed road edge of the target area.

[0053] The various embodiments in this application are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the device and medium embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the description of the method embodiments.

[0054] The devices and media provided in this application are one-to-one with the methods. Therefore, the devices and media also have similar beneficial technical effects as their corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the devices and media will not be repeated here.

[0055] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0056] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0057] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0058] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0059] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0060] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0061] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0062] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0063] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for generating an autonomous driving map, characterized in that, The method includes: The target area for generating the autonomous driving map is determined, and regional images of the target area are collected in different frames using camera sensors installed on the positioning robot. By using a pre-trained image recognition model, road edge curves in the region image are identified, and the starting point and key points in the road edge curves are determined to obtain road edge point cloud data composed of the starting point and the key points. Connect the starting points of different road edge curves to generate the initial road edge corresponding to the target area, and determine whether there are any breaks in the initial road edge; If it exists, add several smoothing points at the breakpoint, and smooth the initial road edge according to the several smoothing points to obtain the processed road edge of the target area; Add several smoothing points at the breakpoint, specifically including: Based on the direction of travel of the positioning robot, the first breakpoint and the last breakpoint of the initial road edge corresponding to the breakpoint are determined; the first breakpoint is the last point on the initial road edge before the breakpoint along the direction of travel, and the last breakpoint is the first point on the initial road edge after the breakpoint along the direction of travel. Two first neighboring points adjacent to the first end breakpoint and two second neighboring points adjacent to the last end breakpoint are determined on the edge of the initial road. Based on the first neighboring points and the second neighboring points, the first tangent line and the second tangent line at the breakpoint are determined accordingly. Determine the intersection point between the first tangent line and the second tangent line, connect the intersection point to the first end breakpoint and the last end breakpoint respectively, and determine the first midpoint of the connecting segment between the intersection point and the first end breakpoint, and the second midpoint of the connecting segment between the intersection point and the last end breakpoint; The first midpoint and the second midpoint are used as smoothing points and added to the breakpoint.

2. The method for generating an autonomous driving map according to claim 1, characterized in that, After identifying road edge curves in the region image, the method further includes: For each frame of a region image, determine the direction of travel of the positioning robot when acquiring the region image; In the case where there are multiple road edge curves in the area image, determine the tangent lines corresponding to the multiple road edge curves and the deviation between the tangent lines and the direction of travel; If the deviation is greater than a preset value, the corresponding road edge curve will be removed from the plurality of road edge curves.

3. The method for generating an autonomous driving map according to claim 1, characterized in that, Determine the starting point and key points in the road edge curve to obtain road edge point cloud data composed of the starting point and the key points, specifically including: For each frame of regional image, the point that is at a preset distance from the starting point of the road edge curve along the traveling direction of the positioning robot is taken as the key point in the road edge curve. According to the direction of travel and the time sequence of the images of the corresponding areas of the road edge curves, the starting point and the key points are written into the point cloud file in sequence to obtain road edge point cloud data composed of the starting points and key points of different road edge curves.

4. The method for generating an autonomous driving map according to claim 3, characterized in that, Before designating a point at a predetermined distance from the starting point of the road edge curve along the travel direction of the positioning robot as a key point in the road edge curve, the method further includes: Obtain the length of the road edge curve corresponding to the current frame region image; If the length of the road edge curve is less than the preset distance, then the points on the road edge curve corresponding to the current frame region image will not be written into the point cloud data.

5. The method for generating an autonomous driving map according to claim 1, characterized in that, The initial road edge is smoothed based on the aforementioned smoothing points to obtain the processed road edge of the target area, specifically including: According to the direction of travel of the positioning robot, a first position sequence is generated, which is composed of the first end breakpoint, the smooth point and the last end breakpoint arranged in sequence; Determine the third midpoints of each connecting segment formed by adjacent points in the first site sequence, and generate a corresponding second site sequence based on the first end breakpoint, each of the third midpoints, and the first end breakpoint; Based on the second site sequence, generate the smooth road edge corresponding to the breakpoint, and determine the inflection point in the smooth road edge; Determine the curvature corresponding to the inflection point. If the curvature is greater than the preset curvature, then determine the fourth midpoint of each segment formed by connecting adjacent points in the second point sequence. By sequentially connecting the first breakpoint, each of the fourth midpoints, and the last breakpoint, the road edge of the target area is obtained.

6. The method for generating an autonomous driving map according to claim 1, characterized in that, After obtaining the processed road edges of the target area, the method further includes: Obtain a point cloud map corresponding to the target area, and determine the actual road edge corresponding to the target area based on the point cloud map; the point cloud map and the autonomous driving map are in the same coordinate system. The road edge of the target area is compared with the actual road edge to determine whether the road edge deviates from the actual road edge. If so, the road edge is corrected.

7. The method for generating an autonomous driving map according to claim 1, characterized in that, After obtaining the processed road edge of the target area, the method further includes: The target area where the positioning robot is located is determined to be a designated area, which is an area where the lane lines are straight and the curves are right-angled bends. The initial direction of the road edge is determined based on the initial distance segment of the road edge; By traversing each edge point on the road edge, the projected distance of each edge point to the initial starting point on the road edge in the initial direction is determined; the initial starting point is the starting point of the road edge at the initial distance segment. If the projection distance of the current edge point is less than the projection distance of the previous edge point, then the current edge point is determined as a turning point, and the turning point is taken as the new initial starting point to obtain the road edge of the straight target area.

8. An autonomous driving map generation device, characterized in that, include: At least one processor; as well as, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the steps of the method as claimed in any one of claims 1-7.

9. A non-volatile computer storage medium storing computer-executable instructions, characterized in that, The computer-executable instructions are configured to perform the steps of the method as described in any one of claims 1-7.

Images