A curb detection method, device, intelligent vehicle, and computer-readable storage medium

Abstract

The present invention provides a curb detection method, device, intelligent vehicle, and computer-readable storage medium, which are applied to an intelligent vehicle. After the obtained point cloud map is processed into a grid; the grid and grid box within the point cloud map where the intelligent vehicle is currently located are obtained, and the curb box is obtained according to the grid box. Topological points are obtained based on the grid box and the curb box. The curb box includes multiple turning points, and the topological points include the vertices of the grid box located within the curb box, the intersection points of the grid box and the curb box, and the direction information of the intersection points. Finally, multiple turning points are connected according to a preset sorting, and the topological points are connected; to output the curb data within the grid. The present invention obtains the grid box in the point cloud map where the intelligent vehicle is currently located, then obtains the curb data within the grid box according to the grid box, and finally determines and outputs the final curb data, thereby reducing the data processed by the intelligent vehicle, accelerating the calculation speed, reducing the memory while improving the accuracy of curb detection.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving, and particularly to a curb detection method, device, intelligent vehicle, and computer-readable storage medium.

Background Art

[0002] When an autonomous driving unmanned vehicle is traveling, it needs to query the curb data around the unmanned vehicle. In the prior art, the curb data is given in the form of multiple simple polygons. When querying the curb data, it is necessary to query the data of all curbs globally each time. Although the required information can be obtained in this way, due to processing too much redundant information, problems such as slow calculation speed and excessive memory overhead are caused.

Summary of the Invention

[0003] Embodiments of the present invention provide a curb segmentation method, device, and autonomous driving vehicle, mainly solving the technical problems of excessive redundancy in processing curb information, slow calculation speed, and large memory overhead in the prior art.

[0004] To solve the above technical problems, a technical solution adopted in an embodiment of the present invention is: to provide a curb detection method, the method including:

[0005] Obtain a point cloud map and perform grid processing on the point cloud map;

[0006] Obtain the grid and grid box where the intelligent vehicle is currently located in the point cloud map, and obtain a curb box within a preset range of the grid. The curb box includes multiple turning points, and the multiple turning points are connected in a preset order;

[0007] Obtain topological points according to the grid box and the curb box. The topological points include the vertices of the grid box located within the curb box, the intersection points of the grid box and the curb box, and the direction information of the intersection points;

[0008] Establish a topological relationship according to the topological points, and connect the topological points according to the topological relationship;

[0009] Output the curb data within the grid.

[0010] Optionally, the obtaining topological points according to the grid box and the curb box includes:

[0011] Obtain the vertices of the grid box;

[0012] Determine the vertices of the grid box located within the curb box as the topological points;

[0013] Calculate the intersection points of the grid box and the curb box in sequence according to the order, and obtain the order of the intersection points and the direction information of the intersection points.

[0014] Optionally, the topological points include at least one first intersection point in the direction of entering the grid box and at least one second intersection point in the direction of exiting the grid box. The step of establishing a topological relationship according to the topological points and connecting the topological points according to the topological relationship specifically includes:

[0015] According to the sorting of the intersection points, obtain the first intersection point and the next adjacent second intersection point, and connect the first intersection point and the next adjacent second intersection point along the curb box;

[0016] Search leftward from the second intersection point along the grid box to find the first first intersection point, record the vertices passing through the grid box, and connect the second intersection point, the first first intersection point, and the vertices of the grid box in sequence;

[0017] Repeat the above steps until the connection lines of the first intersection point, the second intersection point, and the vertices of the grid box form a closed topological loop to establish a topological relationship;

[0018] According to the topological relationship, connect the first intersection point, the second intersection point, the vertices of the grid box, and the turning points between the first intersection point and the second intersection point to form a curb loop.

[0019] Optionally, after the step of establishing a topological relationship according to the topological points and connecting the topological points according to the topological relationship, the method further includes:

[0020] Determine that all the topological points have established the topological relationship.

[0021] Optionally, the method further includes:

[0022] Detect whether the turning point is on the curb box;

[0023] If so, jitter the turning point so that there is a gap between the turning point and the curb box.

[0024] To solve the above technical problems, another technical solution adopted in the embodiments of the present invention is: to provide a curb detection device, the device includes:

[0025] A first acquisition module, configured to acquire a point cloud map and perform grid processing on the point cloud map;

[0026] A second acquisition module, configured to acquire the grid and grid box where the intelligent vehicle is currently located in the point cloud map, and acquire a curb box within a preset range of the grid. The curb box includes a plurality of turning points, and the plurality of turning points are connected according to a preset sorting;

[0027] A third acquisition module, configured to acquire topological points according to the grid frame and the curb frame, where the topological points include vertices of the grid frame located within the curb frame, intersections of the grid frame and the curb frame, and direction information of the intersections;

[0028] A connection module, configured to establish a topological relationship according to the topological points and connect the topological points according to the topological relationship;

[0029] An output module, configured to output curb data within the grid.

[0030] Optionally, the third acquisition module further includes:

[0031] A first acquisition unit, configured to acquire vertices of the grid frame;

[0032] A determination unit, configured to determine vertices of the grid frame located within the curb frame as the topological points;

[0033] A calculation unit, configured to sequentially calculate intersections of the grid frame and the curb frame according to the sorting, and obtain the sorting of the intersections and the direction information of the intersections.

[0034] Optionally, the topological points include at least one first intersection in the direction of entering the grid frame and at least one second intersection in the direction of exiting the grid frame. The connection module specifically includes:

[0035] A first connection unit, configured to obtain the first intersection and the next adjacent second intersection according to the sorting of the intersections, and connect the first intersection and the next adjacent second intersection along the curb frame;

[0036] A search unit, configured to search leftward along the grid frame from the second intersection to find the first intersection, record vertices of the grid frame passed through, and sequentially connect the second intersection, the first first intersection, and the vertices of the grid frame;

[0037] A second connection unit, configured to repeat the above steps until a closed topological loop is formed by the connections of the first intersection, the second intersection, and the vertices of the grid frame, and establish a topological relationship;

[0038] A third connection unit, configured to connect the first intersection, the second intersection, the vertices of the grid frame, and turning points between the first intersection and the second intersection to form a curb loop according to the topological relationship.

[0039] To solve the above technical problems, another technical solution adopted in the embodiments of the present invention is: providing an intelligent vehicle, where the intelligent vehicle includes:

[0040] At least one processor; and,

[0041] A memory communicatively connected to the at least one processor; wherein,

[0042] The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the curb detection method as described above.

[0043] To solve the above technical problems, another technical solution adopted in the embodiments of the present invention is: to provide a computer-readable storage medium storing computer-executable instructions for causing a computer to execute the curb detection method as described above.

[0044] Different from the related art, the present invention provides a curb detection method, device, intelligent vehicle and computer-readable storage medium, which are applied to an intelligent vehicle. The curb detection method, device, intelligent vehicle and computer-readable storage medium obtain a point cloud map and perform grid processing on the point cloud map; then obtain the grid and grid frame where the intelligent vehicle is currently located in the point cloud map, and obtain a curb frame within a preset range of the grid. The curb frame includes a plurality of turning points, and the plurality of turning points are connected in a preset order; obtain topological points according to the grid frame and the curb frame. The topological points include the vertices of the grid frame located within the curb frame, the intersection points of the grid frame and the curb frame, and the direction information of the intersection points; finally, establish a topological relationship according to the topological points and connect the topological points according to the topological relationship; to output the curb data within the grid. The present invention obtains the grid frame where the intelligent vehicle is currently located in the point cloud map, then obtains the curb data within the grid frame according to the grid frame, and finally determines and outputs the final curb data, thereby reducing the data processed by the intelligent vehicle, accelerating the calculation speed, reducing the memory while improving the accuracy of curb detection.

Description of the Drawings

[0045] One or more embodiments are exemplarily illustrated by corresponding drawings. These exemplary illustrations do not constitute limitations on the embodiments. Elements with the same reference numerals in the drawings are represented as similar elements, unless otherwise stated, and the drawings in the drawings do not constitute a proportional limitation.

[0046] Figure 1 is a flowchart of a curb detection method provided by an embodiment of the present invention;

[0047] Figure 2 is a schematic diagram of grid processing of a point cloud map provided by an embodiment of the present invention;

[0048] Figure 3It is to obtain the grid box where the intelligent vehicle is located and the road curb within its preset range;

[0049] Figure 4 It is the detection result diagram of the road curb detection method provided by the embodiment of the present invention;

[0050] Figure 5 It is the structural schematic diagram of a road curb detection device provided by the embodiment of the present invention;

[0051] Figure 6 It is the hardware structure schematic diagram of the intelligent vehicle that executes the above method provided by the embodiment of the present invention.

Specific Embodiment

[0052] In order to make the purpose, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention and are not used to limit the present invention.

[0053] It should be noted that if there is no conflict, the various features in the embodiments of the present invention can be combined with each other, and all are within the protection scope of the present invention. In addition, although the functional modules are divided in the device schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different module division from that in the device schematic diagram or the order in the flowchart.

[0054] Unless otherwise defined, all technical and scientific terms used in this specification have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs. The terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not used to limit the present invention. The term "and / or" used in this specification includes any and all combinations of one or more of the related listed items.

[0055] Please refer to Figure 1 , Figure 1 It is the flowchart of a road curb detection method provided by the embodiment of the present invention, which is applied to an intelligent vehicle. The road curb detection method includes the following steps:

[0056] S01. Obtain a point cloud map and perform grid processing on the point cloud map.

[0057] Specifically, a lidar is installed on the intelligent vehicle, and the lidar is used to obtain a point cloud map. Among them, the lidar refers to a scanning sensor that uses non-contact laser ranging technology. It mainly detects targets by emitting laser beams, and forms a point cloud and obtains data by collecting the reflected beams. These data can be generated into an accurate three-dimensional stereoscopic image after optoelectronic processing.

[0058] Optionally, after obtaining the point cloud map by the intelligent vehicle, where the point cloud map further includes a curb, and the curb refers to an obstacle in the point cloud map. During the driving process of the intelligent vehicle, the lidar will obtain a three-dimensional image of the current road section, and the points of the three-dimensional image form the point cloud map. If the height coordinate in the three-dimensional coordinates of a certain point cloud obtained by the lidar is higher than the height coordinates of other point clouds in the three-dimensional coordinates, then the certain point is the curb in the point cloud map. Then, the obtained point cloud map is subjected to grid processing, such as Figure 2 shown, the point cloud map is divided into multiple rectangular grids, where each Divided cell represents a rectangular grid, object refers to the curb in the point cloud map, and subobject refers to the curb divided after grid processing.

[0059] S02. Obtain the grid and grid frame where the intelligent vehicle is currently located in the point cloud map, and obtain the curb frame within the preset range of the grid. The curb frame includes multiple turning points, and the multiple turning points are connected in a preset order.

[0060] Specifically, the intelligent vehicle travels in the point cloud map. When the intelligent vehicle travels into a certain grid frame, the grid and grid frame of the current position of the intelligent vehicle are obtained. Since the point cloud map includes a curb, after the point cloud map is subjected to grid processing, the grid frame also includes the curb within the preset range of the current grid. As Figure 3 shown, the curb intersects with the current grid frame, that is, a part of the curb is within the grid frame. During the process of curb detection, the intersection of the grid frame and the curb can be regarded as the intersection of a rectangle and a polygon. Inside the rectangular frame, the turning points of the polygon are obtained, and then a starting point is randomly selected, and the turning points are connected in a preset order. The preset order can be connected in the counterclockwise direction, that is, a starting point is randomly selected, and then the next turning point in the counterclockwise connection direction with the starting point is obtained, and the starting point is connected to the turning point, and so on until all the curb frames within the grid frame are obtained. Further, the preset order can also be clockwise sorting.

[0061] Optionally, in this embodiment, by obtaining a single grid frame, then obtaining the curb in the single grid frame, and finally combining the curb situations in all single grid frames together, the data volume of curb detection processing can be reduced, thereby increasing the accuracy of curb detection. Further, in other embodiments, the curb detection method can also be to process two grid frames simultaneously or multiple grid frames simultaneously.

[0062] S03. Obtain topological points based on the grid frame and the curb frame. The topological points include the vertices of the grid frame located within the curb frame, the intersection points between the grid frame and the curb frame, and the direction information of the intersection points.

[0063] Among them, the topology mainly studies the connection relationship between points and lines in an image. Specifically, the physical substance is abstracted into "points" that are independent of the size and shape of the physical substance, and the line segments connecting the physical substances are abstracted into "lines", and then the relationship between these points and lines is represented in the form of a graph.

[0064] Optionally, obtain the vertices of the grid frame; determine the vertices of the grid frame located within the curb frame as the topological points; calculate the intersection points between the grid frame and the curb frame in sequence according to the sorting, and obtain the sorting of the intersection points and the direction information of the intersection points.

[0065] Specifically, first obtain the vertices of the grid frame where the intelligent vehicle is located, then, according to the grid frame and the curb frame, determine whether the vertex is within the curb frame. If the vertex is within the curb frame, determine the vertex within the curb frame as a topological point. Then, connect the vertex to the turning point according to a preset order, and further calculate the intersection points between the curb frame and the grid frame. Among them, the intersection points also include direction information.

[0066] S04. Establish a topological relationship based on the topological points, and connect the topological points according to the topological relationship.

[0067] Optionally, the topological points include at least one first intersection point in the direction of entering the grid frame and at least one second intersection point in the direction of exiting the grid frame. According to the sorting of the intersection points, obtain the first intersection point and the next adjacent second intersection point, and connect the first intersection point and the next adjacent second intersection point along the curb frame; search leftward along the grid frame from the second intersection point to find the first first intersection point, record the vertices of the grid frame passed through, and connect the second intersection point, the first first intersection point, and the vertices of the grid frame in sequence; repeat the above steps until the connection of the first intersection point, the second intersection point, and the vertices of the grid frame forms a closed topological loop to establish a topological relationship; according to the topological relationship, connect the first intersection point, the second intersection point, the vertices of the grid frame, and the turning point between the first intersection point and the second intersection point to form a curb loop.

[0068] Specifically, the topological points include the direction information of the points, for example, input from the outside of the grid box to the inside of the grid box, or output from the inside of the grid box to the outside of the grid box. Then, connect the topological points according to the direction information and position information carried by the topological points. Among them, randomly select a topological point as the starting point, then according to the direction of the topological point, find the adjacent next topological point, and connect the starting point and the next topological point according to the direction of the topological point, and so on, until all topological points and turning points are connected, and finally form a curb loop within the grid box.

[0069] Further, after establishing the topological relationship according to the topological points and connecting the topological points according to the topological relationship, the method further includes determining that all the topological points have established the topological relationship. Specifically, after the curb loop within the grid box is connected, it is also necessary to detect whether there is still an unconnected curb loop within the grid box. If so, it is necessary to re-find the turning points and topological points within the grid box to ensure that all curb loops within the grid box have been found. In the above way, it is avoided that some curb loops are not found due to errors, thereby reducing the accuracy of the curb detection.

[0070] S05. Output the curb data within the grid.

[0071] Please refer to Figure 4 , Figure 4 which is the detection result diagram of the curb detection method provided by the embodiment of the present invention. As Figure 4 shown, if it is determined that all topological relationships in the grid box are established, the curb data within the current grid box is sent to the intelligent vehicle so that the intelligent vehicle can avoid driving along the curb.

[0072] Further, the curb detection method further includes detecting whether the turning point is on the curb frame; if so, jitter the turning point so that there is a gap between the turning point and the curb frame.

[0073] Specifically, after the radar sensor performs grid processing on the obtained point cloud map, the grid box and the curb frame are obtained. Among them, both the grid box and the curb frame are composed of points transformed from the point cloud. Therefore, if there is an intersection point between the grid box and the curb frame, that is, whether the points in the grid box coincide with the points in the curb frame. If they coincide, it is necessary to jitter the intersection point so that there is a gap between the turning point and the curb frame. Further, since the topological points for constructing the curb frame include the direction information of the points, if there is an intersection point between the curb frame and the grid box, it will cause errors in the curb detection, thereby reducing the detection accuracy.

[0074] An embodiment of the present invention provides a curb detection method, which is applied to an intelligent vehicle. The method includes obtaining a point cloud map and performing grid processing on the point cloud map; then obtaining the grid and grid frame where the intelligent vehicle is currently located in the point cloud map, and obtaining a curb frame within a preset range of the grid. The curb frame includes a plurality of turning points, and the plurality of turning points are connected in a preset order; obtaining topological points according to the grid frame and the curb frame, where the topological points include the vertices of the grid frame located within the curb frame, the intersection points of the grid frame and the curb frame, and the direction information of the intersection points; finally, establishing a topological relationship according to the topological points and connecting the topological points according to the topological relationship; to output the curb data within the grid. The present invention obtains the grid frame where the intelligent vehicle is currently located in the point cloud map, then obtains the curb data within the grid frame according to the grid frame, and finally determines and outputs the final curb data, thereby reducing the data processed by the intelligent vehicle, accelerating the calculation speed, reducing the memory, and improving the accuracy of curb detection.

[0075] Please refer to Figure 5 , Figure 5 which is a schematic structural diagram of a curb detection device provided by an embodiment of the present invention. As Figure 5 shown, the curb detection device 40 includes a first acquisition module 41, a second acquisition module 42, a third acquisition module 43, a connection module 44, and an output module 45.

[0076] The first acquisition module 41 is used to acquire a point cloud map and perform grid processing on the point cloud map.

[0077] The second acquisition module 42 is used to acquire the grid and grid frame where the intelligent vehicle is currently located in the point cloud map, and acquire a curb frame within a preset range of the grid. The curb frame includes a plurality of turning points, and the plurality of turning points are connected in a preset order.

[0078] The third acquisition module 43 is used to obtain topological points according to the grid frame and the curb frame. The topological points include the vertices of the grid frame located within the curb frame, the intersection points of the grid frame and the curb frame, and the direction information of the intersection points.

[0079] The third acquisition module 43 includes an acquisition unit 431, a determination unit 432, and a calculation unit 433.

[0080] Specifically, the acquisition unit 431 is used to acquire the vertices of the grid frame;

[0081] The determination unit 432 is used to determine the vertices of the grid frame located within the curb frame as the topological points;

[0082] The calculation unit 433 is configured to calculate the intersections of the grid boxes and the curb boxes in sequence according to the sorting, and obtain the sorting of the intersections and the direction information of the intersections.

[0083] The connection module 44 is configured to establish a topological relationship based on the topological points, and connect the topological points according to the topological relationship.

[0084] Specifically, the connection module 44 includes a first connection unit 441, a search unit 442, a second connection unit 443, and a third connection unit 444.

[0085] The first connection unit 441 is configured to obtain the first intersection and the next adjacent second intersection according to the sorting of the intersections, and connect the first intersection and the next adjacent second intersection along the curb box.

[0086] The search unit 442 is configured to search leftward along the grid box from the second intersection to find the first first intersection, record the vertices of the grid box passed through, and connect the second intersection, the first first intersection, and the vertices of the grid box in sequence.

[0087] The second connection unit 443 is configured to repeat the above steps until the connection of the first intersection, the second intersection, and the vertices of the grid box forms a closed topological loop to establish a topological relationship.

[0088] The third connection unit 444 is configured to connect the first intersection, the second intersection, the vertices of the grid box, and the turning points between the first intersection and the second intersection according to the topological relationship to form a curb loop.

[0089] The output module 45 is configured to output the curb data within the grid.

[0090] It should be noted that the above curb detection device can execute the curb detection method provided by the embodiments of the present invention, and has corresponding functional modules and beneficial effects for executing the method. For technical details not described in detail in the embodiments of the curb detection device, reference can be made to the curb detection method provided by the embodiments of the present invention.

[0091] Please refer to Figure 6 , an intelligent vehicle 30 is provided in an embodiment of the present invention. The intelligent vehicle 30 includes: at least one processor 31, Figure 6 Taking one processor 31 as an example; a memory 32 communicatively connected to the at least one processor 31, Figure 6 Taking the connection through a bus as an example.

[0092] Among them, the memory 32 stores instructions executable by the at least one processor 31. When the instructions are executed by the at least one processor 31, the at least one processor 31 can execute the above-mentioned curb detection method.

[0093] The memory 32, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules, such as the program instructions / modules corresponding to the curb detection method in the embodiments of the present invention. The processor 31 executes various functional applications and data processing of the intelligent vehicle 30 by running the non-volatile software programs, instructions, and modules stored in the memory 32, that is, to implement the curb detection method in the above method embodiments.

[0094] The memory 32 may include a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function. In addition, the memory 32 may include high-speed random access memory and may also include non-volatile memory. For example, it includes at least one disk storage device, a flash memory device, or other non-volatile solid-state storage devices. In some embodiments, the memory 32 may optionally include a memory remotely disposed relative to the processor 31.

[0095] The one or more modules are stored in the memory 32. When executed by the one or more processors 31, they execute the curb detection method in any of the above method embodiments. For example, they execute the Figure 1 method steps described above.

[0096] The intelligent vehicle 30 is also connected to other devices to better execute the method provided in the embodiments of the present invention. For example, it can be electrically connected to a display screen or other displays, and can be remotely communicatively connected to the communication devices of target users, etc., which are not listed one by one here.

[0097] The above intelligent vehicle can execute the method provided in the embodiments of the present invention and has functional modules corresponding to the execution of the method. For technical details not described in detail in this embodiment, reference can be made to the method provided in the embodiments of the present invention.

[0098] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0099] The embodiments of the present application also provide a computer-readable storage medium. The computer-readable storage medium stores computer-executable instructions, which are executed by one or more processors. For example, the method steps described above are executed to implement Figure 1 and the functions of each module in Figure 5 are realized.

[0100] Through the description of the above embodiments, those of ordinary skill in the art can clearly understand that each embodiment can be implemented by means of software plus a general hardware platform, and of course, it can also be implemented by hardware. Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. Among them, the storage medium can be a magnetic disk, an optical disc, a read-only memory (ROM), or a random access memory (RAM), etc.

[0101] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; under the idea of the present invention, the technical features in the above embodiments or different embodiments can also be combined, and the steps can be implemented in any order, and there are many other changes in different aspects of the present invention as described above. For the sake of brevity, they are not provided in detail; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A curb detection method, applied to an intelligent vehicle, characterized in that, the method includes: Obtain a point cloud map, the point cloud map includes curbs, and grid the point cloud map to obtain grids corresponding to the point cloud map, wherein the grids include grid frames and curb frames; Obtain the grid and grid frame where the intelligent vehicle is currently located in the point cloud map, and obtain curb frames within a preset range of the grid. The curb frames include multiple turning points, and the multiple turning points are connected in a preset order. Among them, the curb frame is a polygon; Obtain topological points according to the grid frame and the curb frame. The topological points include the vertices of the grid frame located within the curb frame, the intersection points of the grid frame and the curb frame, and the direction information of the intersection points; Establish a topological relationship according to the topological points, and connect the topological points according to the topological relationship; Output the curb data within the grid; Among them, the topological points include at least one first intersection point in the direction of entering the grid frame and at least one second intersection point in the direction of exiting the grid frame. The step of establishing a topological relationship according to the topological points and connecting the topological points according to the topological relationship specifically includes: According to the sorting of the intersection points, obtain the first intersection point and the next adjacent second intersection point, and connect the first intersection point and the next adjacent second intersection point along the curb frame; Search left along the grid frame from the second intersection point to find the first intersection point, record the vertices of the grid frame passed through, and connect the second intersection point, the first first intersection point, and the vertices of the grid frame in sequence; Repeat the above steps until the connection of the first intersection point, the second intersection point, and the vertices of the grid frame forms a closed topological loop to establish a topological relationship; According to the topological relationship, connect the first intersection point, the second intersection point, the vertices of the grid frame, and the turning points between the first intersection point and the second intersection point to form a curb loop.

2. The method according to claim 1, characterized in that, the obtaining topological points according to the grid frame and the curb frame includes: Obtain the vertices of the grid frame; Determine the vertices of the grid frame located within the curb frame as the topological points; Calculate the intersection points of the grid frame and the curb frame in sequence according to the sorting, and obtain the sorting of the intersection points and the direction information of the intersection points.

3. The method according to any one of claims 1-2, characterized in that, after the step of establishing a topological relationship according to the topological points and connecting the topological points according to the topological relationship, the method further includes: Determine that all the topological points have established the topological relationship.

4. The method according to claim 1, characterized in that, the method further includes: Detect whether the turning point is on the curb frame; If so, jitter the turning point so that there is a gap between the turning point and the curb frame.

5. A curb detection device, applied to an intelligent vehicle, characterized in that, the device includes: A first acquisition module, configured to acquire a point cloud map and grid the point cloud map; A second acquisition module, configured to acquire the grid and the grid frame where the intelligent vehicle is currently located in the point cloud map, and acquire a curb frame within a preset range of the grid, where the curb frame includes a plurality of turning points, and the plurality of turning points are connected in a preset order; A third acquisition module, configured to acquire topological points according to the grid frame and the curb frame, where the topological points include the vertices of the grid frame located within the curb frame, the intersection points of the grid frame and the curb frame, and the direction information of the intersection points; A connection module, configured to establish a topological relationship according to the topological points, and connect the topological points according to the topological relationship; An output module, configured to output the curb data within the grid; Wherein, the topological points include at least one first intersection point in the direction of entering the grid frame and at least one second intersection point in the direction of exiting the grid frame, and the connection module specifically includes: A first connection unit, configured to obtain the first intersection point and the next adjacent second intersection point according to the sorting of the intersection points, and connect the first intersection point and the next adjacent second intersection point along the curb frame; A search unit, configured to search leftward along the grid frame from the second intersection point to find the first intersection point, record the vertices of the grid frame passed through, and connect the second intersection point, the first first intersection point, and the vertices of the grid frame in sequence; A second connection unit, configured to repeat the above steps until the connection lines of the first intersection point, the second intersection point, and the vertices of the grid frame form a closed topological loop, and establish a topological relationship; A third connection unit, configured to connect the first intersection point, the second intersection point, the vertices of the grid frame, and the turning points between the first intersection point and the second intersection point to form a curb loop according to the topological relationship.

6. The apparatus according to claim 5, wherein, the third acquisition module further includes: An acquisition unit, configured to acquire the vertices of the grid frame; A determination unit, configured to determine that the vertices of the grid frame located within the curb frame are the topological points; A calculation unit, configured to calculate the intersection points of the grid frame and the curb frame in sequence according to the sorting, and obtain the sorting of the intersection points and the direction information of the intersection points.

7. An intelligent vehicle, wherein, the intelligent vehicle 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, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method according to any one of claims 1-4.

8. A computer-readable storage medium, wherein, the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to cause a computer to execute the method according to any one of claims 1-4.

Images