Vehicle cruise and cruise mapping method, vehicle-mounted control system, vehicle and medium

By dividing and merging road section information in the vehicle cruise construction map, the problem of insufficient positioning accuracy in multi-floor parking lots is solved, and a more efficient and accurate vehicle cruise map construction is achieved to ensure that the vehicle can enter the target parking space correctly.

CN118936448APending Publication Date: 2024-11-12SHENZHEN DEEPROUTE AI CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410992078.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-23
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The existing vehicle cruise construction technology has insufficient positioning accuracy in multi-floor parking lots, resulting in the vehicle being unable to drive to the finish line correctly.

Method used

By dividing forward and reverse sections of the vehicle's driving sections, and combining semantic information and point cloud information based on the merged sections, a more accurate vehicle cruise map is formed, reducing the road section information with the same effect, and improving the accuracy and efficiency of map construction.

Benefits of technology

It improves the accuracy and efficiency of vehicle cruise construction, ensures that the vehicle can be correctly positioned and entered the target parking space, and improves the safety and experience of autonomous driving.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118936448A_ABST
    Figure CN118936448A_ABST
Patent Text Reader

Abstract

The invention discloses a vehicle cruise and cruise mapping method, a vehicle-mounted control system, a vehicle and a medium. The method comprises the following steps: acquiring a driving road section of a vehicle, and semantic information and point cloud information collected in the driving process of the vehicle; wherein the semantic information at least comprises parking spaces; performing forward road section and reverse road section division on the driving road section, and performing road section combination based on the forward road section and the reverse road section to obtain a combined road section; and based on the combined road section, combining the corresponding semantic information and the point cloud information to form a vehicle cruise map. According to the method, the mapping accuracy and the mapping efficiency can be improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of vehicle cruise technology, and in particular to vehicle cruise and cruise mapping methods, vehicle-mounted control systems, vehicles, and media. Background Art

[0002] Autonomous driving technology is a cutting-edge technology that integrates sensor technology, artificial intelligence, advanced control and other technologies. It is one of the revolutionary advances in the automotive industry in recent years. Today, autonomous driving is no longer limited to structured roads. The automatic cruise function enables the vehicle to automatically memorize and reproduce the driving operation from the garage entrance to the target parking space, completing the self-driving connection from the structured road to the target parking space, and improving the self-driving experience.

[0003] In the related mapping cruise technology, there is still a problem of insufficient accuracy in the cruise mapping of vehicles automatically entering parking spaces. The cruise positioning from the user driving into the garage to reaching the target parking space depends on the registration based on the point cloud map. Inaccurate positioning will cause the vehicle to fail to drive to the destination correctly. For example, in the cruise of multiple floors, the similarity of the upper and lower floors will cause a large deviation between the actual cruise position of the vehicle and the position in the point cloud map for positioning, and the vehicle cannot complete the cruise according to the wrong positioning information. Summary of the invention

[0004] In order to solve the above problems, the vehicle cruise and cruise mapping method, vehicle-mounted control system, vehicle, and medium provided in this application can improve the accuracy and efficiency of mapping.

[0005] In order to solve the above technical problems, a technical solution adopted in the present application is: to provide a vehicle cruise mapping method, the method comprising: obtaining the driving section of the vehicle, and the semantic information and point cloud information collected during the vehicle driving process; wherein the semantic information at least includes the parking space; dividing the driving section into forward sections and reverse sections, and merging the sections based on the forward sections and the reverse sections to obtain merged sections; based on the merged sections, merging the corresponding semantic information and point cloud information to form a vehicle cruise map.

[0006] Among them, the forward section includes a first forward section and a second forward section, the formation time of the first forward section is earlier than the formation time of the second forward section, and the sections are merged based on the forward section and the reverse section to obtain the merged section, including: removing the reverse section, and determining the intersection position between the first forward section and the second forward section; obtaining the first sub-section between the starting trajectory point and the intersection position in the first forward section; obtaining the second sub-section between the intersection position and the ending trajectory point in the second forward section; merging the sections based on the first sub-section and the second sub-section to obtain the merged section.

[0007] Among them, merging the sections based on the first sub-section and the second sub-section to obtain the merged section includes: obtaining trajectory point pairs between the first sub-section and the second sub-section that meet preset conditions; removing the remaining trajectory points between the trajectory point pairs; and performing trajectory smoothing between the trajectory point pairs to merge the first sub-section and the second sub-section to obtain the merged section.

[0008] Among them, obtaining a trajectory point pair between the first sub-segment and the second sub-segment that meets the preset conditions includes: taking the intersection position as the starting point, obtaining a first target trajectory point from the first sub-segment, and obtaining a second target trajectory point from the second sub-segment; in response to the distance between the first target trajectory point and the second target trajectory point being less than the preset distance, and the angle between the line between the first target trajectory point and the second target trajectory point and the tangent of the first target trajectory point being less than the preset angle, determining the first target trajectory point and the second target trajectory point as a trajectory point pair.

[0009] After obtaining the driving section of the vehicle, the method further includes: using the driving section of the vehicle to divide the section attributes; wherein the section attributes include at least one of an uphill section, a downhill section, a curved road section and a straight road section.

[0010] Among them, based on the merged road section, the corresponding semantic information and point cloud information are merged to form a vehicle cruising map, including: obtaining the semantic information of all speed bumps and parking spaces in the merged road section, and tracking different parking spaces and speed bumps to form map semantic information belonging to the merged road section; performing voxel filtering on the single-frame point cloud information in the merged road section to obtain an occupied grid point cloud sub-map; and forming a vehicle cruising map based on the merged road section, map semantic information and occupied grid point cloud sub-map.

[0011] Among them, target tracking is performed on different parking spaces and speed bumps to form map semantic information belonging to the merged road section, including: converting parking spaces and speed bumps into rectangles; tracking different parking spaces and speed bumps to obtain multiple observation information; using multiple observation information, adding point-to-plane constraints and opposite-side parallel constraints from each rectangular corner point to the plane formed by the other three corner points of the rectangle, and using nonlinear optimization to obtain the optimal position of parking spaces and speed bumps in the map.

[0012] To solve the above technical problems, another technical solution adopted in the present application is: to provide a vehicle cruise mapping method, the method comprising: in response to a vehicle entering a cruiseable area, starting a pre-mapping mode, and obtaining a first cruise map according to the method provided by any of the above technical solutions; in response to an external input mapping instruction, starting a formal mapping mode, and obtaining a second cruise map according to the method provided by any of the above technical solutions; and combining the first cruise map and the second cruise map to form a vehicle cruise map.

[0013] Among them, the cruising area is a multi-story parking area.

[0014] To solve the above technical problems, another technical solution adopted in the present application is: providing a vehicle cruising method, the method comprising: obtaining a cruising map in response to a cruising instruction; wherein the cruising map is established by using the method provided by any of the above technical solutions; automatically forming a cruising trajectory according to the cruising map and the target parking space, and controlling the vehicle to enter the target parking space according to the cruising trajectory.

[0015] To solve the above technical problems, another technical solution adopted in the present application is: to provide a vehicle-mounted control system, which includes a processor and a memory coupled to the processor; wherein the memory is used to store computer programs, and the processor is used to execute computer programs to implement a method provided by any of the above technical solutions.

[0016] In order to solve the above technical problems, another technical solution adopted in the present application is: providing a vehicle, the vehicle includes a vehicle-mounted control system provided by the above technical solution.

[0017] To solve the above technical problems, another technical solution adopted in the present application is: to provide a computer-readable storage medium, which is used to store a computer program. When the computer program is executed by a processor, it is used to implement the method provided by any of the above technical solutions.

[0018] The beneficial effects of the embodiments of the present application are as follows: different from the prior art, the vehicle cruise and cruise mapping method, vehicle-mounted control system, vehicle, and medium provided by the present application divide the driving section into forward sections and reverse sections during the vehicle cruise mapping process, and merge the sections based on the forward sections and the reverse sections to obtain merged sections, thereby reducing the section information with the same function and improving the subsequent mapping accuracy and efficiency; further, based on the merged sections, the corresponding semantic information and point cloud information are merged to improve the correlation between the semantic information and the point cloud information and the sections, thereby forming a more accurate vehicle cruise map. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0020] Figure 1 It is a flow chart of an embodiment of a vehicle cruise mapping method provided by the present application;

[0021] Figure 2It is a flow chart of another embodiment of the vehicle cruise mapping method provided by the present application;

[0022] Figure 3 and Figure 4 It is a schematic diagram of the application scenario provided by this application;

[0023] Figure 5 It is a flowchart of an embodiment of step 25 provided by the present application;

[0024] Figure 6 , Figure 7 It is a schematic diagram of the application scenario provided by this application;

[0025] Figure 8 It is a flow chart of another embodiment of the vehicle cruise mapping method provided by the present application;

[0026] Fig. 9 It is a flow chart of another embodiment of the vehicle cruise mapping method provided by the present application;

[0027] Fig.10 It is a flow chart of an embodiment of a vehicle cruising method provided by the present application;

[0028] Fig.11 It is a structural schematic diagram of an embodiment of a vehicle-mounted control system provided by the present application;

[0029] Fig.12 It is a structural schematic diagram of an embodiment of a vehicle provided by the present application;

[0030] Fig.13 It is a structural schematic diagram of an embodiment of a computer-readable storage medium provided by the present application. DETAILED DESCRIPTION

[0031] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It will be understood that the specific embodiments described herein are only used to explain the present application, rather than to limit the present application. It should also be noted that, for ease of description, only some but not all structures related to the present application are shown in the drawings. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present application.

[0032] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0033] Autonomous driving technology is a cutting-edge technology that integrates sensor technology, artificial intelligence, advanced control and other technologies. It is one of the revolutionary advances in the automotive industry in recent years. Today, autonomous driving is no longer limited to structured roads. The automatic cruise function enables the vehicle to automatically memorize and reproduce the driving operation from the garage entrance to the target parking space, completing the self-driving connection from the structured road to the target parking space, and improving the self-driving experience.

[0034] In the related art, there is still a problem of insufficient accuracy in cruise mapping for vehicles to automatically enter parking spaces.

[0035] Based on this, the present application proposes that in the process of vehicle cruise mapping, the driving section is divided into forward sections and reverse sections, and the sections are merged based on the forward sections and the reverse sections to obtain merged sections, so as to reduce the section information with the same function and improve the subsequent mapping accuracy and efficiency; further, based on the merged sections, the corresponding semantic information and point cloud information are merged to improve the correlation between the semantic information and the point cloud information and the section, thereby forming a more accurate vehicle cruise map. The following is an explanation of the section information with the same function: If the functions of two section information are the same, the information of one section can be obtained by deleting one of them or merging the two. If the two driving sections correspond to the same target, the two driving sections can be merged into one driving section. When the section information is used later, the information of one driving section can be used to reduce the subsequent algorithm process.

[0036] Please refer to any of the following embodiments for details.

[0037] See also Figure 1 , Figure 1 This is a flow chart of an embodiment of a vehicle cruise mapping method provided by the present application. The method includes:

[0038] Step 11: Obtain the driving section of the vehicle, as well as the semantic information and point cloud information collected during the driving process of the vehicle; wherein the semantic information at least includes the parking space.

[0039] In some embodiments, the road traveled by the vehicle can be divided into multiple sub-roads with forward and reverse directions, straight roads, curves, flat roads, and ramps in real time according to the curvature of the road, the road slope, and the driving direction of the current vehicle.

[0040] In some embodiments, the road segment on which the vehicle is traveling may be used to divide the road segment attributes; wherein the road segment attributes include at least one of an uphill road segment, a downhill road segment, a curved road segment, and a straight road segment.

[0041] In some embodiments, the image acquisition device on the vehicle can be used to acquire images during driving, and then the corresponding deep learning network can be used to identify the semantic information in the acquired images, such as acquiring the semantic information of all speed bumps and parking spaces in the road section.

[0042] In some embodiments, the laser radar equipment on the vehicle can be used to collect point cloud information while the vehicle is driving, so as to form a corresponding point cloud map.

[0043] Step 12: Divide the driving section into forward sections and reverse sections, and merge the sections based on the forward sections and the reverse sections to obtain a merged section.

[0044] In some embodiments, taking the cruising area as a parking lot as an example, the forward section may refer to the section where the vehicle drives toward the parking space, and the reverse section may refer to the section where the vehicle drives away from the parking space.

[0045] During the cruise mapping process, although the vehicle can move back and forth in the cruise area, the cruise mapping is intended to construct the best parking route. At this time, the road sections where the vehicle travels in the cruise area are actually redundant and erroneous. The reverse sections will affect the mapping, so the reverse sections can be deleted. Then, several forward sections are merged to obtain the merged sections. In this way, the best cruise mapping path is determined.

[0046] Step 13: Based on the merged road sections, the corresponding semantic information and point cloud information are merged to form a vehicle cruising map.

[0047] In some embodiments, as the vehicle is driving, each road section that the vehicle travels on will generate corresponding semantic information and point cloud information. Therefore, after the road sections are merged, the corresponding semantic information and point cloud information can be merged according to the merged road sections, so as to reduce the data volume of the semantic information and point cloud information and reduce storage pressure.

[0048] In this embodiment, during the vehicle cruise mapping process, the driving section is divided into forward sections and reverse sections, and the sections are merged based on the forward sections and the reverse sections to obtain merged sections, thereby reducing the section information with the same function and improving the subsequent mapping accuracy and efficiency; further, based on the merged sections, the corresponding semantic information and point cloud information are merged to improve the correlation between the semantic information and the point cloud information and the sections, thereby forming a more accurate vehicle cruise map.

[0049] See also Figure 2 , Figure 2 This is a flow chart of another embodiment of the vehicle cruise mapping method provided by the present application. The method includes:

[0050] Step 21: Obtain the driving section of the vehicle, as well as the semantic information and point cloud information collected during the driving of the vehicle; wherein the semantic information at least includes the parking space.

[0051] Step 22: Divide the driving section into a forward section and a reverse section, remove the reverse section, and determine the intersection position between the first forward section and the second forward section.

[0052] In this embodiment, the forward section at least includes a first forward section and a second forward section. The first forward section is formed earlier than the second forward section.

[0053] Combination Figure 3 and Figure 4 To explain:

[0054] like Figure 3 As shown, the formal section is divided into a first forward section link1, a reverse section link2 and a second forward section link3. The first forward section link1 is formed earlier than the reverse section link2, and the reverse section link2 is formed earlier than the second forward section link3.

[0055] When merging, remove the reverse link link2, then Figure 4 There are only the first forward link link1 and the second forward link link3. Then the intersection position between the first forward link link1 and the second forward link link3 is determined.

[0056] Step 23: Obtain the first sub-segment between the starting trajectory point and the intersection position in the first forward segment.

[0057] In some embodiments, since the first positive section and the second positive section are both positive sections, and the formation time of the first positive section is earlier than the formation time of the second positive section, it is explained that the section from the starting track point to the intersection position in the first positive section can also be used as the corresponding section of the second positive section. Therefore, in the merging process, the first sub-section between the starting track point and the intersection position in the first positive section is obtained.

[0058] Step 24: Obtain a second sub-segment between the intersection position and the end trajectory point in the second forward segment.

[0059] In some embodiments, since the first positive segment and the second positive segment are both positive segments, and the formation time of the first positive segment is earlier than the formation time of the second positive segment, it is indicated that the segment from the intersection position to the end track point in the second positive segment can also be used as the corresponding segment of the first positive segment. Therefore, during the merging process, the second sub-segment between the intersection position and the end track point in the second positive segment is obtained.

[0060] It can be understood that, since the first positive road section and the second positive road section are both positive road sections, both road sections may be generated when the vehicle drives toward the parking space, and therefore the two road sections can be combined to form a merged road section.

[0061] Step 25: Merge the road segments based on the first sub-segment and the second sub-segment to obtain a merged road segment.

[0062] In some embodiments, the first sub-segment and the second sub-segment may be directly merged and connected at the intersection to form a merged segment.

[0063] In some embodiments, see Figure 5 , step 25 may be the following process:

[0064] Step 251: Obtain a pair of trajectory points between the first sub-segment and the second sub-segment that meet preset conditions.

[0065] In some embodiments, the driving section of the vehicle is essentially composed of continuous trajectory points. Therefore, when merging based on the intersection position, it is also necessary to consider the smoothing of the first sub-segment and the second sub-segment when merging. For example, although both are forward sections, the angle corresponding to the intersection position of the two sections is close to 90 degrees, then the merged section is a section with a right angle section, which will increase the difficulty of vehicle cruising, so smoothing is required.

[0066] In some embodiments, starting from the intersection position, a first target trajectory point is obtained from the first sub-segment, and a second target trajectory point is obtained from the second sub-segment. In response to the distance between the first target trajectory point and the second target trajectory point being less than a preset distance, and the angle between the connecting line between the first target trajectory point and the second target trajectory point and the tangent of the first target trajectory point being less than a preset angle, the first target trajectory point and the second target trajectory point are determined as a trajectory point pair. In some embodiments, the preset distance can be 1 meter, 2 meters, 3 meters, 5 meters or 10 meters, etc., which can be set according to actual conditions. The preset angle can be 20 degrees, 25 degrees, 30 degrees or 35 degrees, etc., which can be set according to actual conditions.

[0067] Since the first sub-segment is in the forward direction, the corresponding trajectory point is obtained in the first sub-segment in the reverse direction from the intersection position as the starting point, as the first target trajectory point. Since the second sub-segment is in the forward direction, the corresponding trajectory point is obtained in the second sub-segment in the forward direction from the intersection position as the starting point, as the second target trajectory point. When the first pair of the first target trajectory point and the second target trajectory point does not meet the requirements, a new trajectory point is obtained from the first sub-segment as the first target trajectory point, and the comparison is performed again, or a new trajectory point is obtained from the second sub-segment as the second target trajectory point, and the comparison is performed again, and the iteration is performed until a trajectory point pair that meets the preset conditions is found between the first sub-segment and the second sub-segment.

[0068] In summary, the determination logic of trajectory point pairs can be as follows:

[0069] Since the first sub-segment is in the forward direction, the corresponding trajectory points are obtained in the first sub-segment in the reverse direction from the intersection position as the starting point, and the first trajectory point set is formed in the order of acquisition. Since the second sub-segment is in the forward direction, the corresponding trajectory points are obtained in the second sub-segment in the forward direction from the intersection position as the starting point, and the second trajectory point set is formed in the order of acquisition.

[0070] By using an alternating traversal method, the trajectory points between the first trajectory point set and the second trajectory point set are compared to find a trajectory point pair that meets the preset conditions.

[0071] Specifically, the alternating traversal method is as follows:

[0072] The first trajectory point is obtained from the first trajectory point set, and the trajectory point is used to compare with all the trajectory points in the second trajectory point set. If there is a target trajectory point that meets the requirements in the second trajectory point set, the first trajectory point and the first target trajectory point that meets the requirements are taken as a trajectory point pair that meets the preset conditions.

[0073] If there is no target trajectory point that meets the requirements in the second trajectory point set, the first trajectory point is removed from the first trajectory point set, and the first trajectory point is obtained from the second trajectory point set. This trajectory point is compared with all trajectory points in the first trajectory point set. If there is a target trajectory point that meets the requirements in the first trajectory point set, the first trajectory point and the first target trajectory point that meets the requirements are taken as a trajectory point pair that meets the preset conditions.

[0074] If there is no target trajectory point that meets the requirements in the first trajectory point set, the first trajectory point is removed from the second trajectory point set, and a new first trajectory point is obtained from the first trajectory point set. This trajectory point is compared with all the trajectory points in the second trajectory point set. If there is a target trajectory point that meets the requirements in the second trajectory point set, the first trajectory point and the first target trajectory point that meets the requirements are taken as a trajectory point pair that meets the preset conditions.

[0075] This cycle is repeated until a pair of trajectory points between the first sub-segment and the second sub-segment that meets the preset conditions is obtained.

[0076] Step 252: Remove the remaining trajectory points between the trajectory point pairs.

[0077] After the trajectory point pairs are determined, there may be trajectory points between the trajectory point pairs that did not meet the requirements before, so the remaining trajectory points between the trajectory point pairs can be removed.

[0078] Step 253: Perform trajectory smoothing between trajectory point pairs to merge the first sub-segment and the second sub-segment to obtain a merged segment.

[0079] In some embodiments, a B-spline curve may be used to smooth the trajectory between the trajectory point pairs, and corresponding trajectory points may be added between the trajectory point pairs, thereby merging the first sub-segment and the second sub-segment to obtain a merged segment.

[0080] In some embodiments, in combination Figure 6 and Figure 7 To explain:

[0081] like Figure 6 As shown, the first sub-segment link11 includes trajectory points d, e, f, and g. The second sub-segment link31 includes trajectory points a, b, and c. Starting from the intersection position, trajectory point d is obtained from the first sub-segment, and trajectory point a is obtained from the second sub-segment. Then the distance between trajectory point a and trajectory point d is calculated, and the angle between the line connecting trajectory point a and trajectory point d and the tangent line of trajectory point a is calculated.

[0082] In response to the distance between trajectory point a and trajectory point d being less than a preset distance, and the angle between the connecting line between trajectory point a and trajectory point d and the tangent line of trajectory point a being less than a preset angle, trajectory point a and trajectory point d are determined as a trajectory point pair.

[0083] If the distance between trajectory point a and trajectory point d and / or the angle between the line connecting trajectory point a and trajectory point d and the tangent line of trajectory point a do not meet the requirements, a new trajectory point is searched again.

[0084] Taking the case where the distance between trajectory point a and trajectory point d and / or the angle between the line connecting trajectory point a and trajectory point d and the tangent line of trajectory point a do not meet the requirements, trajectory point e is obtained from the first sub-segment and trajectory point a is obtained from the second sub-segment. Then, the distance between trajectory point e and trajectory point a is calculated, and the angle between the line connecting trajectory point e and trajectory point a and the tangent line of trajectory point d is calculated.

[0085] In response to the distance between trajectory point e and trajectory point a being less than a preset distance, and the angle between the connecting line between trajectory point e and trajectory point a and the tangent line of trajectory point d being less than a preset angle, trajectory point e and trajectory point a are determined as a trajectory point pair.

[0086] Then, the remaining trajectory points between trajectory point e and trajectory point a are removed, such as trajectory point d.

[0087] Then, the trajectory is smoothed between the trajectory point e and the trajectory point a to merge the first sub-segment and the second sub-segment to obtain a merged segment. Figure 7 The merged road segment shown has trajectory points e, f, g, a, b, c, and a newly added trajectory point between trajectory point e and trajectory point a.

[0088] It can be understood that there may be multiple forward sections and reverse sections, and the forward sections and reverse sections of each area may be post-processed in the above manner.

[0089] Step 26: Based on the merged road section, the corresponding semantic information and point cloud information are merged to form a vehicle cruising map.

[0090] In this embodiment, during the vehicle cruise mapping process, the driving section is divided into forward sections and reverse sections, and the sections are merged based on the forward sections and the reverse sections to obtain merged sections, thereby reducing the section information with the same function and improving the subsequent mapping accuracy and efficiency; further, based on the merged sections, the corresponding semantic information and point cloud information are merged to improve the correlation between the semantic information and the point cloud information and the sections, thereby forming a more accurate vehicle cruise map.

[0091] Furthermore, by merging the forward section and the reverse section, redundant trajectory points are deleted and trajectory smoothing is performed during the merging process, thereby improving the driving comfort of the final merged section.

[0092] See also Figure 8 , Figure 8 1 is a flow chart of another embodiment of the vehicle cruise mapping method provided by the present application. The method includes:

[0093] Step 81: Acquire the driving section of the vehicle, as well as the semantic information and point cloud information collected during the driving of the vehicle; wherein the semantic information at least includes the parking space.

[0094] Step 82: Divide the driving section into forward sections and reverse sections, and merge sections based on the forward sections and the reverse sections to obtain merged sections.

[0095] Steps 81 to 82 are the same or similar technical solutions as any embodiment of the present application and are not described in detail here.

[0096] Step 83: Acquire semantic information of all speed bumps and parking spaces in the merged road section, and perform target tracking on different parking spaces and speed bumps to form map semantic information belonging to the merged road section.

[0097] In some embodiments, parking spaces and speed bumps are converted into rectangles; different parking spaces and speed bumps are tracked to obtain multiple observation information; using the multiple observation information, point-to-plane constraints and opposite-side parallel constraints are added from each corner point of the rectangle to the plane formed by the other three corner points of the rectangle, and nonlinear optimization is used to obtain the optimal position of the parking spaces and speed bumps in the map, thereby improving the positioning accuracy of subsequent cruise maps.

[0098] Step 84: voxel filtering is performed on the single-frame point cloud information in the merged road section to obtain an occupied grid point cloud sub-image.

[0099] In some embodiments, point cloud information is downsampled and stored using an incremental voxel filter for global positioning during cruising.

[0100] Step 85: Form a vehicle cruising map based on the merged road segments, map semantic information and occupied grid point cloud sub-map.

[0101] In some embodiments, the merged road segments, map semantic information and occupied grid point cloud sub-graphs may be stored separately. When using a vehicle cruise map, the merged road segments, map semantic information and occupied grid point cloud sub-graphs are obtained to quickly construct a vehicle cruise map.

[0102] In this embodiment, during the vehicle cruise mapping process, the driving section is divided into forward sections and reverse sections, and the sections are merged based on the forward sections and the reverse sections to obtain merged sections, thereby reducing the section information with the same function and improving the subsequent mapping accuracy and efficiency; further, based on the merged sections, the corresponding semantic information and point cloud information are merged to improve the correlation between the semantic information and the point cloud information and the sections, thereby forming a more accurate vehicle cruise map.

[0103] Furthermore, the single-frame point cloud information in the merged road section is subjected to voxel filtering to obtain an occupied grid point cloud sub-image, thereby reducing the storage resource occupation of the point cloud information.

[0104] In one application scenario, after the vehicle starts cruising and mapping, the roads that the vehicle has traveled are accurately divided and post-processed. The curvature, slope and driving direction of the road that the vehicle is currently traveling on are obtained by using the current position and current posture information of the vehicle in real time. The collected driving route is divided into different sections using various information of different sections, so as to divide different floors and the upper limit of the cruising speed of different sections. At the same time, the trajectory post-processing can be performed on the forward and reverse sections to generate the available cruising trajectory of the vehicle.

[0105] After determining the corresponding road, the semantic information of all speed bumps, parking spaces, etc. in the road section is obtained from the perception feature map, and different parking spaces and speed bumps are tracked and optimized to form the map semantic information of the road section. Speed ​​bump information can be used to inform the vehicle of the speed bump location, so that the vehicle can slow down in advance when passing the speed bump during cruising, increasing the passing comfort of the vehicle; parking space semantic information can be used for the vehicle driver to park in the selected parking space during cruising, so that parking has greater freedom.

[0106] After determining the corresponding road, the cached continuous single-frame point cloud with high memory usage is stored as an occupancy grid point cloud sub-map with small memory usage using an incremental voxel filter, which greatly reduces the system memory pressure. This point cloud sub-map is used for vehicle relocation and positioning update during cruising. At the same time, since the point cloud sub-map is bound to the road section, it can provide timely feedback to the vehicle driver on whether the vehicle has entered the cruiseable area and whether the existing map needs to be updated. Among them, the occupancy grid point cloud sub-map uses a hash table for random search to accelerate the insertion process of the point cloud, and incrementally samples the point cloud, so that it has the excellent characteristics of fast speed and small memory usage in the generation of the occupancy grid point cloud sub-map.

[0107] In some embodiments, the memorized route can be divided into different sections using various information of different sections, and the floor attribute of the section from the ramp to the next ramp can be determined by whether there is an uphill or downhill attribute. The sections can be divided into curves and straights using the attributes, and different cruising speed limits can be provided for different sections in combination with the ramp attributes to optimize the cruising riding experience.

[0108] In some embodiments, the forward and reverse attributes of each road segment can be used to post-process the road segment with reverse driving. Link1, link2 and link3 are sequential road segments, where link1 and link3 are forward road segments and link2 is a reverse road segment. In the road segment post-processing, first remove the reverse road segment link2; find the intersection between link1 and link3, and cut off the first forward road segment link1 from the intersection, leaving only the starting part, and cut off the second forward road segment link3 from the intersection, leaving only the end part; find the trajectory point pairs belonging to the front and rear forward road segments, so that the distance between the point pairs is within a certain threshold, and the angle α between the tangent direction of the point belonging to the front link in the point pair and the direction of the point pair connection line is less than a certain threshold; connect the point pairs, remove redundant points, and use B-spline curves to smooth the trajectory, so as to obtain the merged road segment that can be used downstream. After the road segments are merged, the semantic elements generated by the semantic information construction module and the point cloud subgraph generated by the point cloud map construction module are merged at the same time.

[0109] In the actual cruise mapping process, after the vehicle drives to the cruiseable area, the current indoor and outdoor environment is detected in real time using deep learning algorithms and GNSS information. When the vehicle is detected to enter the indoor area from the outdoor area, the mapping system automatically starts to build the map without feeling. When the driver clicks the button to start memorizing the cruise route, the mapping system can continue to build the map normally. After the mapping is completed, the point cloud map stored in the mapping process can be matched with the lidar point cloud when the vehicle enters the garage again, so as to determine the location of the vehicle in the garage earlier and avoid safety and experience problems caused by the positioning system mislocating the floor in the multi-story garage.

[0110] See also Fig. 9 , Fig. 9 1 is a flow chart of another embodiment of the vehicle cruise mapping method provided by the present application. The method includes:

[0111] Step 91: In response to the vehicle entering the cruiseable area, the pre-map mode is turned on to obtain a first cruise map.

[0112] In some embodiments, after the pre-mapping mode is turned on, mapping is performed according to the technical solution of any of the above embodiments to obtain a first cruise map.

[0113] Step 92: In response to an external input mapping instruction, a formal mapping mode is started to obtain a second cruise map.

[0114] In some embodiments, after the formal mapping mode is turned on, mapping is performed according to the technical solution of any of the above embodiments to obtain a second cruise map.

[0115] Step 93: Combine the first cruise map and the second cruise map to form a vehicle cruise map.

[0116] In some embodiments, the navigable area is a multi-story parking area.

[0117] In the process of cruise mapping, it is actually the vehicle driver who actively triggers the cruise mapping function. In more scenarios, the vehicle driver will start the cruise mapping only after reaching the designated parking floor, so it is easy to have positioning errors. Based on this, the cruise mapping scheme proposed in this application can be divided into two parts: pre-mapping and formal mapping according to the timing when the vehicle driver triggers the formal mapping. The vehicle control system can identify the environmental image through the collected environmental image, that is, according to the deep learning perception information, it can identify the current indoor and outdoor environment, and at the same time, combined with the GNSS information, it can judge the timing of the vehicle from the outdoors to the indoors. When the vehicle drives indoors, the vehicle control system turns on the pre-mapping mode and starts pre-mapping without feeling. When the vehicle driver clicks the cruise mapping learning button, the vehicle control system switches to the formal mapping mode, and then combines the information of the pre-mapping mode and the formal mapping mode to obtain the entire cruise map from the entrance of the parking area to the required parking floor. This method can effectively avoid the situation where the wrong floor is located during cruising due to cross-floor mapping, ensuring the cruising safety and cruising experience of the vehicle driver.

[0118] See also Fig.10 , Fig.10 1 is a flow chart of an embodiment of a vehicle cruising method provided by the present application. The method comprises:

[0119] Step 101: In response to a cruise instruction, a cruise map is acquired.

[0120] Among them, the cruise map is established using the method provided by any of the above technical solutions.

[0121] Step 102: automatically forming a cruise trajectory according to the cruise map and the target parking space, and controlling the vehicle to enter the target parking space according to the cruise trajectory.

[0122] In some embodiments, the cruise map will display parking spaces that can be parked. After the vehicle driver selects the target parking space, the corresponding cruise track will be automatically generated according to the cruise map, and the vehicle will be controlled to enter the target parking space according to the cruise track. It can be understood that the parking spaces displayed in the cruise map are several parking spaces collected through semantic recognition during the cruise mapping process.

[0123] In some embodiments, if the target parking space does not have a corresponding cruise trajectory, a corresponding cruise trajectory can be automatically planned for the target parking space based on the existing cruise sections in the cruise map.

[0124] See also Fig.11 , Fig.111 is a schematic diagram of the structure of an embodiment of a vehicle control system provided by the present application. The vehicle control system 110 includes a processor 111 and a memory 112 coupled to the processor 111; wherein the memory 112 is used to store a computer program, and the processor 111 is used to execute the computer program to implement the following method:

[0125] Acquire the driving section of the vehicle, as well as the semantic information and point cloud information collected during the driving process of the vehicle; wherein the semantic information at least includes the parking space; divide the driving section into forward sections and reverse sections, and merge the sections based on the forward sections and the reverse sections to obtain a merged section; based on the merged section, merge the corresponding semantic information and point cloud information to form a vehicle cruising map.

[0126] It can be understood that the processor 111 is also used to execute a computer program to implement the method of any of the above embodiments, which will not be elaborated here.

[0127] See also Fig.12 , Fig.12 1 is a schematic diagram of the structure of an embodiment of a vehicle provided by the present application. The vehicle 120 includes an onboard control system 110. The onboard control system 110 is the onboard control system 110 provided by the above technical solution.

[0128] See also Fig.13 , Fig.13 1 is a schematic diagram of a computer readable storage medium according to an embodiment of the present application. The computer readable storage medium 130 is used to store a computer program 131. When the computer program 131 is executed by a processor, it is used to implement the following method:

[0129] Acquire the driving section of the vehicle, as well as the semantic information and point cloud information collected during the driving process of the vehicle; wherein the semantic information at least includes the parking space; divide the driving section into forward sections and reverse sections, and merge the sections based on the forward sections and the reverse sections to obtain a merged section; based on the merged section, merge the corresponding semantic information and point cloud information to form a vehicle cruising map.

[0130] It can be understood that when the computer program 131 is executed by the processor, it is also used to implement the method of any of the above embodiments, which will not be elaborated here.

[0131] In summary, the vehicle cruise and cruise mapping method, on-board control system, vehicle, and medium provided by the present application divide the driving section into forward sections and reverse sections during the vehicle cruise mapping process, and merge the sections based on the forward sections and the reverse sections to obtain merged sections, thereby reducing the section information with the same function and improving the subsequent mapping accuracy and efficiency; further, based on the merged sections, the corresponding semantic information and point cloud information are merged to improve the correlation between the semantic information and the point cloud information and the sections, thereby forming a more accurate vehicle cruise map.

[0132] In the several embodiments provided in this application, it should be understood that the disclosed methods and devices can be implemented in other ways. For example, the device implementation described above is only illustrative, for example, the division of the circuit or unit is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0133] 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 distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0134] In addition, each functional unit in each embodiment of the present application may be integrated into a first processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0135] The above description is only an implementation method of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made according to the description and drawings of the present application, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A vehicle cruising mapping method, It is characterized in that The method comprises: Acquire the driving section of the vehicle, as well as the semantic information and point cloud information collected during the driving of the vehicle; wherein the semantic information at least includes a parking space; Dividing the driving section into a forward section and a reverse section, and merging sections based on the forward section and the reverse section to obtain a merged section; Based on the merged road section, the corresponding semantic information and point cloud information are merged to form a vehicle cruising map.

2. The method according to claim 1, It is characterized in that The forward segment includes a first forward segment and a second forward segment, the first forward segment is formed earlier than the second forward segment, and the segment merging based on the forward segment and the reverse segment to obtain the merged segment includes: removing the reverse road segment, and determining an intersection position between the first forward road segment and the second forward road segment; Acquire a first sub-segment between the starting trajectory point and the intersection position in the first forward segment; Acquire a second sub-segment between the intersection position and the end trajectory point in the second forward segment; The first sub-segment and the second sub-segment are used to merge the segments to obtain a merged segment.

3. The method according to claim 2, It is characterized in that The step of merging the road segments based on the first sub-segment and the second sub-segment to obtain a merged road segment includes: Acquire a pair of trajectory points between the first sub-segment and the second sub-segment that meet preset conditions; removing the remaining trajectory points between the trajectory point pairs; Trajectory smoothing is performed between the trajectory point pairs to merge the first sub-segment and the second sub-segment to obtain a merged segment.

4. The method according to claim 3, It is characterized in that The obtaining of a track point pair between the first sub-segment and the second sub-segment that meets a preset condition includes: Taking the intersection position as a starting point, obtaining a first target trajectory point from the first sub-segment, and obtaining a second target trajectory point from the second sub-segment; In response to the distance between the first target trajectory point and the second target trajectory point being less than a preset distance, and the angle between a line between the first target trajectory point and the second target trajectory point and a tangent line of the first target trajectory point being less than a preset angle, the first target trajectory point and the second target trajectory point are determined as the trajectory point pair.

5. The method according to claim 1, It is characterized in that After obtaining the driving section of the vehicle, the method further includes: The road segment of the vehicle is used to divide the road segment attributes; wherein the road segment attributes include at least one of an uphill road segment, a downhill road segment, a curved road segment and a straight road segment.

6. The method according to claim 1, It is characterized in that The method of combining corresponding semantic information and point cloud information based on the combined road section to form a vehicle cruising map includes: Acquire semantic information of all speed bumps and parking spaces in the merged road section, and perform target tracking on different parking spaces and speed bumps to form map semantic information belonging to the merged road section; Perform voxel filtering on the single-frame point cloud information in the merged road section to obtain an occupied grid point cloud sub-image; A vehicle cruising map is formed based on the merged road segments, map semantic information and the occupied grid point cloud sub-map.

7. The method according to claim 6, It is characterized in that The target tracking of different parking spaces and speed bumps to form map semantic information belonging to the merged road section includes: Convert the parking spaces and speed bumps into rectangles; Track different parking spaces and speed bumps to obtain multiple observation information; Using the multiple observation information, point-to-plane constraints and opposite-side parallel constraints are added from each rectangular corner point to the plane formed by the other three corner points of the rectangle, and nonlinear optimization is used to obtain the optimal positions of the parking spaces and speed bumps in the map.

8. A vehicle cruising mapping method, It is characterized in that The method comprises: In response to the vehicle entering a cruiseable area, starting a pre-map mode, and obtaining a first cruise map according to the method of any one of claims 1 to 7; In response to an external input mapping instruction, a formal mapping mode is started, and a second cruise map is obtained according to the method of any one of claims 1 to 7; The first cruising map and the second cruising map are combined to form a vehicle cruising map.

9. The method according to claim 8, It is characterized in that The cruiseable area is a multi-storey parking area.

10. A vehicle cruising method, It is characterized in that The method comprises: In response to a cruise instruction, a cruise map is obtained; wherein the cruise map is established by using the method according to any one of claims 1 to 9; A cruising track is automatically formed according to the cruising map and the target parking space, and the vehicle is controlled to enter the target parking space according to the cruising track.

11. A vehicle-mounted control system, It is characterized in that The vehicle-mounted control system includes a processor and a memory coupled to the processor; wherein the memory is used to store a computer program, and the processor is used to execute the computer program to implement the method as described in any one of claims 1-10.

12. A vehicle, It is characterized in that The vehicle includes the on-board control system of claim 11.

13. A computer-readable storage medium, It is characterized in that The computer-readable storage medium is used to store a computer program, and when the computer program is executed by a processor, the computer program is used to implement the method according to any one of claims 1 to 10.