Multi-vehicle cooperative laser mapping method and device, terminal equipment and storage medium

Abstract

The invention discloses a multi-vehicle cooperative laser mapping method and device, terminal equipment and a storage medium. The method comprises the following steps: acquiring a first three-dimensional point cloud map corresponding to a plurality of vehicles; wherein each first three-dimensional point cloud map corresponds to a vehicle; each vehicle generates a first three-dimensional point cloud map in different areas of a to-be-mapped area; determining a to-be-deleted map in all the first three-dimensional point cloud maps; deleting the map to be deleted to obtain a second three-dimensional point cloud map; and according to the positioning data and the pose data of each second three-dimensional point cloud map, carrying out map splicing to complete the construction of the three-dimensional point cloud map of the region to be mapped. According to the invention, the map deletion operation and the map splicing operation are carried out on the point cloud data collected by a plurality of vehicles together, the region to be mapped can be divided into a plurality of small regions, synchronous data collection of the three-dimensional point clouds of the plurality of small regions is carried out, and the three-dimensional mapping efficiency is greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of three-dimensional mapping technology, and in particular to a multi-vehicle collaborative laser mapping method, device, terminal equipment and storage medium. Background Art

[0002] The SLAM algorithm uses the acquired point cloud data to incrementally construct maps based on the perception of the environment by sensors such as lidar, and achieves real-time positioning based on the registration of the map and the point cloud and the odometer. Most of today's laser mapping solutions use single-vehicle SLAM for mapping, which requires a single vehicle to run through the entire area before mapping can be built. This will lead to a significant increase in the running time and cumulative mileage of the SLAM algorithm, thereby increasing the cumulative error of the SLAM algorithm; and single-vehicle mapping sometimes requires passing the same section of road multiple times, recording a large amount of redundant data, affecting the quality of the generated point cloud map and reducing the efficiency of mapping.

[0003] Therefore, a laser mapping strategy is urgently needed to solve the problem of low efficiency of three-dimensional mapping. Summary of the invention

[0004] Embodiments of the present invention provide a multi-vehicle collaborative laser mapping method, apparatus, terminal device, and storage medium to solve the problem of low efficiency in three-dimensional mapping.

[0005] In order to solve the above problems, an embodiment of the present invention provides a laser mapping method for multi-vehicle collaboration, including:

[0006] Acquire first three-dimensional point cloud maps corresponding to a plurality of vehicles; wherein each first three-dimensional point cloud map corresponds to a vehicle; and each of the vehicles generates a first three-dimensional point cloud map in different areas of the area to be mapped;

[0007] Among all the first three-dimensional point cloud maps, determine the map to be deleted;

[0008] Deleting the map to be deleted to obtain a second three-dimensional point cloud map;

[0009] Map stitching is performed based on the positioning data and posture data of each of the second three-dimensional point cloud maps to complete the construction of the three-dimensional point cloud map of the area to be mapped.

[0010] As an improvement of the above solution, the step of obtaining a first three-dimensional point cloud map corresponding to a plurality of vehicles includes:

[0011] Obtaining factor information transmitted by each vehicle;

[0012] Based on each factor information, a first three-dimensional point cloud map is generated.

[0013] As an improvement of the above solution, the map stitching is performed according to the positioning data and the posture data of each of the second three-dimensional point cloud maps to complete the construction of the three-dimensional point cloud map of the area to be mapped, including:

[0014] According to the positioning data and the posture data of each of the second three-dimensional point cloud maps, the map stitching operation is repeatedly performed until the number of the second three-dimensional point cloud maps reaches one, thereby completing the construction of the three-dimensional point cloud map of the area to be mapped.

[0015] The map stitching operation includes:

[0016] In the second three-dimensional point cloud map, selecting a preset number of target second three-dimensional point cloud maps;

[0017] In all the second three-dimensional point cloud maps of the target, based on the positioning data and the posture data, the overlapping areas are identified and the overlapping areas are merged to obtain a third three-dimensional point cloud map;

[0018] Using the third three-dimensional point cloud map as the second three-dimensional point cloud map required for the next map stitching operation, and determining whether the number of the current second three-dimensional point cloud maps is one;

[0019] If yes, output the current second three-dimensional point cloud map data to complete the construction of the three-dimensional point cloud map of the area to be mapped;

[0020] If not, re-execute the map stitching operation.

[0021] As an improvement of the above solution, the method of confirming the overlapping area based on the positioning data and the posture data, and merging the overlapping area to obtain a third three-dimensional point cloud map includes:

[0022] According to the positioning data, a second three-dimensional point cloud map of the target with the same positioning data is selected as a three-dimensional point cloud map of the same area;

[0023] In all three-dimensional point cloud maps of the same area, point cloud registration is performed according to the position and posture data, and the area whose registration degree reaches the registration degree threshold is determined as the overlapping area; wherein the overlapping area includes: a first overlapping area and a second overlapping area; wherein the first overlapping area is the overlapping area with the largest loading time; and the second overlapping area is the overlapping area with the non-largest loading time;

[0024] The first overlapping area is retained, and the second overlapping area is deleted;

[0025] Taking the first overlapping area as a connection point, the three-dimensional point cloud map of the same area with the second overlapping area deleted is merged with the three-dimensional point cloud map of the same area with the first overlapping area retained to obtain a third three-dimensional point cloud map.

[0026] As an improvement of the above solution, determining the map to be deleted from all the first three-dimensional point cloud maps includes:

[0027] In the first three-dimensional point cloud map, in response to the user's operation, the point cloud map selected by the user is used as a deletion map;

[0028] Or, in the first three-dimensional point cloud map, according to a preset point cloud database, a point cloud map whose matching degree with the point cloud database is greater than a matching degree threshold is used as a deletion map; wherein the point cloud database is a point cloud map entered in advance by a user.

[0029] As an improvement to the above solution, the factor information is obtained by the vehicle in the following manner:

[0030] The laser point cloud data, IMU data and GPS data collected by the vehicle in the area to be mapped; wherein the laser point cloud data, the IMU data and the GPS data are collected at the same time;

[0031] The vehicle repeatedly performs a mapping operation based on the laser point cloud data, the IMU data, and the GPS data until all the laser point cloud data are completely mapped, thereby obtaining a first three-dimensional point cloud map;

[0032] The mapping operation is specifically as follows:

[0033] Selecting a laser point cloud data from the laser point cloud data on which the mapping operation has not been performed as a target laser point cloud;

[0034] Based on the IMU data corresponding to the target laser point cloud and the current IMU odometer, the target laser point cloud is subjected to motion distortion correction processing to obtain a corrected laser point cloud;

[0035] According to the curvature of the corrected laser point cloud, the type of feature points of the corrected laser point cloud is determined, and the pose data is obtained based on the type of feature points;

[0036] Based on the pose corresponding to the current target laser point cloud and the previous target laser point cloud, the pose change degree is calculated, and the key frame point cloud and LIDAR odometer are confirmed according to the pose change degree;

[0037] And based on the key frame point cloud and the current target laser point cloud, a stitching map and factor information corresponding to the stitching map are obtained;

[0038] Determine whether all laser point cloud data have been mapped;

[0039] If yes, the factor information corresponding to the current spliced ​​map is output;

[0040] If not, the LIDAR odometer is pre-integrated to obtain a new IMU odometer, and the new IMU odometer is used as the IMU odometer for the next mapping operation, and the next mapping operation is performed.

[0041] As an improvement of the above scheme, the key frame point cloud includes: a first key frame point cloud and a second key frame point cloud; the LIDAR odometer includes: a first LIDAR odometer and a second LIDAR odometer; the method of confirming the key frame point cloud and the LIDAR odometer according to the degree of change of posture includes:

[0042] The vehicle makes a judgment based on the degree of change of the posture;

[0043] If the degree of change of the posture is greater than the degree of change threshold and there is a previous target laser point cloud to perform a mapping operation, the vehicle uses the current target laser point cloud as the first initial key frame point cloud and generates an odometer factor at the same time; the vehicle determines the closed loop factor based on closed loop detection; the vehicle obtains the first GPS factor by converting the GPS data corresponding to the target laser point cloud; the vehicle optimizes the first initial key frame point cloud based on the odometer factor, the closed loop factor and the first GPS factor to obtain the first key frame point cloud, and determines the first LIDAR odometer based on the first key frame point cloud;

[0044] If the posture change is less than the change threshold, or there is no previous target laser point cloud to perform mapping operation, the vehicle uses the current target laser point cloud as the second initial key frame point cloud, converts the GPS data corresponding to the target laser point cloud, and obtains the second GPS factor; the vehicle optimizes the second initial key frame based on the second GPS factor, obtains the second key frame point cloud, and determines the second LIDAR odometer based on the second key frame point cloud.

[0045] Accordingly, an embodiment of the present invention further provides a multi-vehicle collaborative laser mapping device, comprising: a data acquisition module, a map selection module, a map deletion module, and a map splicing module;

[0046] The data acquisition module is used to acquire first three-dimensional point cloud maps corresponding to a plurality of vehicles; wherein each first three-dimensional point cloud map corresponds to a vehicle; and each of the vehicles generates a first three-dimensional point cloud map in different areas of the area to be mapped;

[0047] The map selection module is used to determine the map to be deleted from all the first three-dimensional point cloud maps;

[0048] The map deletion module is used to delete the map to be deleted to obtain a second three-dimensional point cloud map;

[0049] The map stitching module is used to perform map stitching based on the positioning data and posture data of each of the second three-dimensional point cloud maps to complete the construction of the three-dimensional point cloud map of the area to be mapped.

[0050] Correspondingly, an embodiment of the present invention further provides a computer terminal device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, a multi-vehicle collaborative laser mapping method as described in the present invention is implemented.

[0051] Correspondingly, an embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute a multi-vehicle collaborative laser mapping method as described in the present invention.

[0052] As can be seen from the above, the present invention has the following beneficial effects:

[0053] The present invention provides a laser mapping method for multi-vehicle collaboration, which obtains a first three-dimensional point cloud map corresponding to several vehicles; wherein each first three-dimensional point cloud map corresponds to a vehicle; each of the vehicles generates a first three-dimensional point cloud map in different areas of the area to be mapped; among all the first three-dimensional point cloud maps, a map to be deleted is determined; the map to be deleted is deleted to obtain a second three-dimensional point cloud map; and map splicing is performed according to the positioning data and posture data of each of the second three-dimensional point cloud maps to complete the construction of the three-dimensional point cloud map of the area to be mapped. The present invention can split the area to be mapped into multiple small areas by performing map deletion and map splicing operations on the point cloud data collected by multiple vehicles, and perform synchronous data collection of three-dimensional point clouds of multiple small areas. Compared with the prior art that can only perform three-dimensional point cloud collection by a single vehicle, the present invention can perform three-dimensional point cloud collection of multiple vehicles by deleting and splicing three-dimensional point cloud maps, which greatly improves the efficiency of three-dimensional mapping. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a flowchart of a laser mapping method for multi-vehicle collaboration provided by an embodiment of the present invention;

[0055] Figure 2 is a schematic diagram of the structure of a multi-vehicle collaborative laser mapping device provided by an embodiment of the present invention;

[0056] Figure 3 This is a schematic diagram of the structure of a terminal device provided by an embodiment of the present invention;

[0057] Figure 4 is a schematic diagram of a mapping control method provided by an embodiment of the present invention;

[0058] Figure 5 It is a schematic diagram of a mapping control method provided by another embodiment of the present invention. DETAILED DESCRIPTION

[0059] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0060] Embodiment 1

[0061] See also Figure 1 , Figure 1 FIG. 1 is a flow chart of a laser mapping method for multi-vehicle collaboration provided by an embodiment of the present invention. Figure 1 As shown, this embodiment includes steps 101 to 104, and each step is specifically as follows:

[0062] Step 101: Obtain first three-dimensional point cloud maps corresponding to a number of vehicles; wherein each first three-dimensional point cloud map corresponds to a vehicle; and each of the vehicles generates a first three-dimensional point cloud map in different areas of the area to be mapped.

[0063] In this embodiment, obtaining a first three-dimensional point cloud map corresponding to a plurality of vehicles includes:

[0064] Obtaining factor information transmitted by each vehicle;

[0065] Based on each factor information, a first three-dimensional point cloud map is generated.

[0066] In a specific embodiment, by receiving the factor information transmitted by the vehicle, a first three-dimensional point cloud map is generated based on the loading function. The loading function mainly reads the factor information retained when the vehicle is mapped, reconstructs the factor graph using each factor information, quickly restores the key frame pose, and then uses the key frame pose and its corresponding point cloud to splice and generate a rough point cloud map and display it on the visualization interface, thereby obtaining the first three-dimensional point cloud map.

[0067] In this embodiment, map stitching is performed based on the positioning data and posture data of each of the second three-dimensional point cloud maps to complete the construction of the three-dimensional point cloud map of the area to be mapped.

[0068] In this embodiment, the factor information is obtained by the vehicle in the following manner:

[0069] The laser point cloud data, IMU data and GPS data collected by the vehicle in the area to be mapped; wherein the laser point cloud data, the IMU data and the GPS data are collected at the same time;

[0070] The vehicle repeatedly performs a mapping operation based on the laser point cloud data, the IMU data, and the GPS data until all the laser point cloud data are completely mapped, thereby obtaining a first three-dimensional point cloud map;

[0071] The mapping operation is specifically as follows:

[0072] Selecting a laser point cloud data from the laser point cloud data on which the mapping operation has not been performed as a target laser point cloud;

[0073] Based on the IMU data corresponding to the target laser point cloud and the current IMU odometer, the target laser point cloud is subjected to motion distortion correction processing to obtain a corrected laser point cloud;

[0074] According to the curvature of the corrected laser point cloud, the type of feature points of the corrected laser point cloud is determined, and the pose data is obtained based on the type of feature points;

[0075] Based on the pose corresponding to the current target laser point cloud and the previous target laser point cloud, the pose change degree is calculated, and the key frame point cloud and LIDAR odometer are confirmed according to the pose change degree;

[0076] And based on the key frame point cloud and the current target laser point cloud, a stitching map and factor information corresponding to the stitching map are obtained;

[0077] Determine whether all laser point cloud data have been mapped;

[0078] If yes, the factor information corresponding to the current spliced ​​map is output;

[0079] If not, the LIDAR odometer is pre-integrated to obtain a new IMU odometer, and the new IMU odometer is used as the IMU odometer for the next mapping operation, and the next mapping operation is performed.

[0080] In a specific embodiment, the vehicle mapping step mainly includes control, recording and building Figure 3 By default, only the control module is turned on, which is responsible for receiving user instructions and turning on or off the recording and mapping functions as required. The control module also monitors the execution status of the recording and mapping functions and feeds back to the user, and writes it into the log file. The specific structure is as follows Figure 4 shown.

[0081] After the recording function is turned on, the program will automatically store the point cloud data, imu (inertial measurement unit) data and GPS data of the area where the vehicle passes. Every 10 frames of point cloud data are read, they are serialized into binary data and stored in the database. When the recording is ended, the remaining unprocessed data will be serialized and written to the database.

[0082] After turning on the map building function, the program will open the database generated when recording the package and read data from it. First, use the IMU and IMU odometer data to estimate the vehicle's transformed posture at the moment of each laser point, and obtain the posture transformation matrix relative to the starting laser point of each frame, so as to perform motion distortion correction on the laser point cloud data and correct the laser point coordinate error caused by vehicle motion. Then, by calculating the curvature, the laser point cloud is divided into edge and plane feature points, and the edge and plane points are respectively aligned with the edge and plane points of the constructed local map to solve the coordinate transformation and motion relationship between the two adjacent frames, so as to obtain the posture of the current frame. Compare the posture of the current frame with the key frame of the previous frame. If a large posture change occurs, generate the next key frame and odometer factor. If there is no key frame of the previous frame, wait for the GPS factor to be generated, and generate the starting key frame based on the GPS factor. The closed loop detection uses the current vehicle's posture to search for historical key frames within a certain range, find the historical key frame closest to the current vehicle, and splice it into a local map with it as the center to align with the current point cloud. If the similarity is high, their posture transformation matrix is ​​calculated to generate a closed loop factor. GPS data needs to be converted from longitude and latitude coordinates to UTM coordinates and can only be used after noise judgment. The generation of a new GPS factor requires a certain distance from the previous GPS factor.

[0083] It should be noted that the laser point cloud sends a frame every 100ms, and the timestamp of the laser point in the frame varies from 0 to 99ms. Therefore, the laser points in the frame are divided into 100 groups, each corresponding to the same timestamp. The angular velocity integral in the IMU data is used to calculate the angle change (roll, pitch, yaw) of the vehicle within 100ms, and the position change (x, y, z) of the vehicle within 100ms is calculated by the position change of the IMU odometer, and then the change is allocated according to the time ratio. For example, if the change of (x, y, z, roll, pitch, yaw) in 100ms is (0.25, 0, 0, 0, 0, 0.1), then the corresponding 10ms posture transformation matrix is ​​(0.025, 0, 0, 0, 0.01).

[0084] It can be understood that the motion distortion correction process is specifically as follows: according to the obtained posture transformation matrix (x, y, z, roll, pitch, yaw), the coordinates (x, y, z) of the laser point are first rotated and then translated.

[0085] It should be noted that the curvature is calculated based on the distance between the laser point and the radar. The square of the distance difference between the five points before and after the current point on the same line is accumulated. If the value is greater than 1.0, it is a corner point, otherwise it is a surface point.

[0086] In the factor graph, the present invention abstracts the key frames into nodes, abstracts each factor into a constraint edge, performs graph optimization, obtains the optimized key frames, and generates the LIDAR odometer based on this. Then, based on the LIDAR odometer, the IMU data is pre-integrated according to the timestamp to obtain the IMU odometer. In order to reduce memory usage, the program will continuously write the various factors and point clouds corresponding to the old key frames into the database during the mapping process, and only retain the most recently used parts in the memory. When the mapping is completed, the remaining parts in the memory will also be saved to the database for later reproduction of the mapping information. After the mapping is completed, a point cloud map, trajectory file and parameter file will be generated. The point cloud map is a group of small squares with a length and width of 30 meters, which are obtained by cutting the complete point cloud map. The trajectory file is generated using key frames to reflect the movement of the vehicle during the mapping process. The parameter file is generated based on the parameters of the vehicle itself. The specific process of the mapping function is as follows Figure 5 shown.

[0087] In a specific embodiment, the key frame contains point cloud corner point and surface point information, as well as the current vehicle posture (x, y, z, roll, pitch, yaw), and the vehicle posture (x, y, z, roll, pitch, yaw) is used to directly generate the lidar odometer (x, y, z, roll, pitch, yaw).

[0088] In this embodiment, the key frame point cloud includes: a first key frame point cloud and a second key frame point cloud; the LIDAR odometer includes: a first LIDAR odometer and a second LIDAR odometer; the step of confirming the key frame point cloud and the LIDAR odometer according to the degree of change of posture includes:

[0089] The vehicle makes a judgment based on the degree of change of the posture;

[0090] If the degree of change of the posture is greater than the degree of change threshold and there is a previous target laser point cloud to perform a mapping operation, the vehicle uses the current target laser point cloud as the first initial key frame point cloud and generates an odometer factor at the same time; the vehicle determines the closed loop factor based on closed loop detection; the vehicle obtains the first GPS factor by converting the GPS data corresponding to the target laser point cloud; the vehicle optimizes the first initial key frame point cloud based on the odometer factor, the closed loop factor and the first GPS factor to obtain the first key frame point cloud, and determines the first LIDAR odometer based on the first key frame point cloud;

[0091] If the posture change is less than the change threshold, or there is no previous target laser point cloud to perform mapping operation, the vehicle uses the current target laser point cloud as the second initial key frame point cloud, converts the GPS data corresponding to the target laser point cloud, and obtains the second GPS factor; the vehicle optimizes the second initial key frame based on the second GPS factor, obtains the second key frame point cloud, and determines the second LIDAR odometer based on the second key frame point cloud.

[0092] Step 102: Determine a map to be deleted from among all first three-dimensional point cloud maps.

[0093] In this embodiment, determining the map to be deleted from all first three-dimensional point cloud maps includes:

[0094] In the first three-dimensional point cloud map, in response to the user's operation, the point cloud map selected by the user is used as a deletion map;

[0095] Or, in the first three-dimensional point cloud map, according to a preset point cloud database, a point cloud map whose matching degree with the point cloud database is greater than a matching degree threshold is used as a deletion map; wherein the point cloud database is a point cloud map entered in advance by a user.

[0096] Step 103: Delete the map to be deleted to obtain a second three-dimensional point cloud map.

[0097] In this embodiment, a map to be deleted is selected, and a deletion operation is performed to remove the map to be deleted from the first three-dimensional point cloud map. There are two ways to select the part to be deleted, one is manual selection, and the other is automatic selection through the play package (i.e., the point cloud map pre-entered by the user in the point cloud database of the present invention).

[0098] Step 104: Map stitching is performed based on the positioning data and the posture data of each of the second three-dimensional point cloud maps to complete the construction of the three-dimensional point cloud map of the area to be mapped.

[0099] In this embodiment, the map stitching is performed according to the positioning data and the posture data of each of the second three-dimensional point cloud maps to complete the construction of the three-dimensional point cloud map of the area to be mapped, including:

[0100] According to the positioning data and the posture data of each of the second three-dimensional point cloud maps, the map stitching operation is repeatedly performed until the number of the second three-dimensional point cloud maps reaches one, thereby completing the construction of the three-dimensional point cloud map of the area to be mapped.

[0101] The map stitching operation includes:

[0102] In the second three-dimensional point cloud map, selecting a preset number of target second three-dimensional point cloud maps;

[0103] In all the second three-dimensional point cloud maps of the target, based on the positioning data and the posture data, the overlapping areas are identified and the overlapping areas are merged to obtain a third three-dimensional point cloud map;

[0104] Using the third three-dimensional point cloud map as the second three-dimensional point cloud map required for the next map stitching operation, and determining whether the number of the current second three-dimensional point cloud maps is one;

[0105] If yes, output the current second three-dimensional point cloud map data to complete the construction of the three-dimensional point cloud map of the area to be mapped;

[0106] If not, re-execute the map stitching operation.

[0107] In this embodiment, the method of confirming the overlapping area based on the positioning data and the posture data, and merging the overlapping area to obtain a third three-dimensional point cloud map includes:

[0108] According to the positioning data, a second three-dimensional point cloud map of the target with the same positioning data is selected as a three-dimensional point cloud map of the same area;

[0109] In all three-dimensional point cloud maps of the same area, point cloud registration is performed according to the position and posture data, and the area whose registration degree reaches the registration degree threshold is determined as the overlapping area; wherein the overlapping area includes: a first overlapping area and a second overlapping area; wherein the first overlapping area is the overlapping area with the largest loading time; and the second overlapping area is the overlapping area with the non-largest loading time;

[0110] The first overlapping area is retained, and the second overlapping area is deleted;

[0111] Taking the first overlapping area as a connection point, the three-dimensional point cloud map of the same area with the second overlapping area deleted is merged with the three-dimensional point cloud map of the same area with the first overlapping area retained to obtain a third three-dimensional point cloud map.

[0112] In a specific embodiment, the second three-dimensional point cloud map to be spliced ​​is first loaded out, and the GPS base points (i.e., the positioning data described in the present invention) set by them are checked to be consistent, so as to ensure that the second three-dimensional point cloud map to be spliced ​​is in the same local area; then, the similar positions in the two maps are found according to the posture data, and the point cloud registration is used to determine whether it belongs to the overlapping area. If it belongs to the overlapping area, a new loop factor is generated and added to the factor graph. For the overlapping area, the point cloud information of the point cloud map B (the first overlapping area) loaded later will be used, and the part of the point cloud information loaded first into the point cloud map A (the second overlapping area) will be removed, so that if the new point cloud map is loaded later, the update of the point cloud information of the overlapping road section can be realized. The spliced ​​point cloud map is also stored in the factor information library, but the parameter information of each vehicle will be saved, and the parameter information will correspond to the point cloud area one by one for subsequent use. If the hardware conditions support it, multiple maps can be spliced ​​at the same time, such as opening 4 processes to splice 8 (i.e., the preset number described in the present invention) point cloud maps at the same time.

[0113] See also Figure 2 , Figure 2 2 is a schematic diagram of the structure of a multi-vehicle collaborative laser mapping device provided by an embodiment of the present invention, comprising: a data acquisition module 201, a map selection module 202, a map deletion module 203 and a map splicing module 204;

[0114] The data acquisition module is used to acquire first three-dimensional point cloud maps corresponding to a plurality of vehicles; wherein each first three-dimensional point cloud map corresponds to a vehicle; and each of the vehicles generates a first three-dimensional point cloud map in different areas of the area to be mapped;

[0115] The map selection module is used to determine the map to be deleted from all the first three-dimensional point cloud maps;

[0116] The map deletion module is used to delete the map to be deleted to obtain a second three-dimensional point cloud map;

[0117] The map stitching module is used to perform map stitching based on the positioning data and posture data of each of the second three-dimensional point cloud maps to complete the construction of the three-dimensional point cloud map of the area to be mapped.

[0118] It can be understood that the above-mentioned system item embodiments correspond to the method item embodiments of the present invention, and can implement the multi-vehicle collaborative laser mapping method provided by any of the above-mentioned method item embodiments of the present invention.

[0119] This embodiment obtains a first three-dimensional point cloud map corresponding to a number of vehicles; wherein each first three-dimensional point cloud map corresponds to a vehicle; each of the vehicles generates a first three-dimensional point cloud map in different areas of the area to be mapped; among all the first three-dimensional point cloud maps, a map to be deleted is determined; the map to be deleted is deleted to obtain a second three-dimensional point cloud map; and map splicing is performed according to the positioning data and posture data of each of the second three-dimensional point cloud maps to complete the construction of the three-dimensional point cloud map of the area to be mapped. The present invention can split the area to be mapped into multiple small areas by performing map deletion and map splicing operations on the point cloud data collected by multiple vehicles, and perform synchronous data collection of three-dimensional point clouds of multiple small areas. Compared with the prior art that can only perform three-dimensional point cloud collection by a single vehicle, the present invention can perform three-dimensional point cloud collection of multiple vehicles by deleting and splicing three-dimensional point cloud maps, which greatly improves the efficiency of three-dimensional mapping.

[0120] Embodiment 2

[0121] See also Figure 3 , Figure 3 It is a schematic diagram of the structure of a terminal device provided in one embodiment of the present invention.

[0122] A terminal device of this embodiment includes: a processor 301, a memory 302, and a computer program stored in the memory 302 and executable on the processor 301. When the processor 301 executes the computer program, the steps of the above-mentioned laser mapping method for multi-vehicle collaboration in the embodiment are implemented, for example Figure 1 Alternatively, when the processor executes the computer program, the functions of each module in the above-mentioned device embodiments are implemented, for example: Figure 2 All modules of the multi-vehicle collaborative laser mapping device are shown.

[0123] In addition, an embodiment of the present invention further provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the multi-vehicle collaborative laser mapping method as described in any of the above embodiments.

[0124] Those skilled in the art will understand that the schematic diagram is merely an example of a terminal device and does not constitute a limitation on the terminal device. The terminal device may include more or fewer components than shown in the diagram, or a combination of certain components, or different components. For example, the terminal device may also include input and output devices, network access devices, buses, etc.

[0125] The processor 301 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor 301 is the control center of the terminal device, and uses various interfaces and lines to connect various parts of the entire terminal device.

[0126] The memory 302 can be used to store the computer program and / or module. The processor 301 implements various functions of the terminal device by running or executing the computer program and / or module stored in the memory and calling the data stored in the memory 302. The memory 302 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0127] Wherein, if the module / unit integrated in the terminal device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0128] It should be noted that the device embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the accompanying drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art may understand and implement it without paying any creative effort.

[0129] The above is a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A laser mapping method for multi-vehicle collaboration, characterized in that: include: Acquire first three-dimensional point cloud maps corresponding to a plurality of vehicles; wherein each first three-dimensional point cloud map corresponds to a vehicle; and each of the vehicles generates a first three-dimensional point cloud map in different areas of the area to be mapped; Among all the first three-dimensional point cloud maps, determine the map to be deleted; Deleting the map to be deleted to obtain a second three-dimensional point cloud map; Map stitching is performed based on the positioning data and posture data of each of the second three-dimensional point cloud maps to complete the construction of the three-dimensional point cloud map of the area to be mapped.

2. The laser mapping method for multi-vehicle collaboration according to claim 1, characterized in that: The obtaining of a first three-dimensional point cloud map corresponding to a plurality of vehicles includes: Obtaining factor information transmitted by each vehicle; Based on each factor information, a first three-dimensional point cloud map is generated.

3. The laser mapping method for multi-vehicle collaboration according to claim 2 is characterized in that: The method of performing map stitching according to the positioning data and the posture data of each of the second three-dimensional point cloud maps to complete the construction of the three-dimensional point cloud map of the area to be mapped includes: According to the positioning data and the posture data of each of the second three-dimensional point cloud maps, the map stitching operation is repeatedly performed until the number of the second three-dimensional point cloud maps reaches one, thereby completing the construction of the three-dimensional point cloud map of the area to be mapped. The map stitching operation includes: In the second three-dimensional point cloud map, selecting a preset number of target second three-dimensional point cloud maps; In all the second three-dimensional point cloud maps of the target, based on the positioning data and the posture data, the overlapping areas are identified and the overlapping areas are merged to obtain a third three-dimensional point cloud map; Using the third three-dimensional point cloud map as the second three-dimensional point cloud map required for the next map stitching operation, and determining whether the number of the current second three-dimensional point cloud maps is one; If yes, output the current second three-dimensional point cloud map data to complete the construction of the three-dimensional point cloud map of the area to be mapped; If not, re-execute the map stitching operation.

4. The laser mapping method for multi-vehicle collaboration according to claim 3 is characterized in that: The method of confirming the overlapping area based on the positioning data and the posture data and merging the overlapping area to obtain a third three-dimensional point cloud map includes: According to the positioning data, a second three-dimensional point cloud map of the target with the same positioning data is selected as a three-dimensional point cloud map of the same area; In all three-dimensional point cloud maps of the same area, point cloud registration is performed according to the position and posture data, and the area whose registration degree reaches the registration degree threshold is determined as the overlapping area; wherein the overlapping area includes: a first overlapping area and a second overlapping area; wherein the first overlapping area is the overlapping area with the largest loading time; and the second overlapping area is the overlapping area with the non-largest loading time; The first overlapping area is retained, and the second overlapping area is deleted; Taking the first overlapping area as a connection point, the three-dimensional point cloud map of the same area with the second overlapping area deleted is merged with the three-dimensional point cloud map of the same area with the first overlapping area retained to obtain a third three-dimensional point cloud map.

5. The laser mapping method for multi-vehicle collaboration according to claim 1, characterized in that: The step of determining a map to be deleted from among all first three-dimensional point cloud maps includes: In the first three-dimensional point cloud map, in response to the user's operation, the point cloud map selected by the user is used as a deletion map; Or, in the first three-dimensional point cloud map, according to a preset point cloud database, a point cloud map whose matching degree with the point cloud database is greater than a matching degree threshold is used as a deletion map; wherein the point cloud database is a point cloud map entered in advance by a user.

6. The laser mapping method for multi-vehicle collaboration according to claim 1, characterized in that: The factor information is obtained by the vehicle in the following manner: The laser point cloud data, IMU data and GPS data collected by the vehicle in the area to be mapped; wherein the laser point cloud data, the IMU data and the GPS data are collected at the same time; The vehicle repeatedly performs a mapping operation based on the laser point cloud data, the IMU data, and the GPS data until all the laser point cloud data are completely mapped, thereby obtaining a first three-dimensional point cloud map; The mapping operation is specifically as follows: Selecting a laser point cloud data from the laser point cloud data on which the mapping operation has not been performed as a target laser point cloud; Based on the IMU data corresponding to the target laser point cloud and the current IMU odometer, the target laser point cloud is subjected to motion distortion correction processing to obtain a corrected laser point cloud; According to the curvature of the corrected laser point cloud, the type of feature points of the corrected laser point cloud is determined, and the pose data is obtained based on the type of feature points; Based on the pose corresponding to the current target laser point cloud and the previous target laser point cloud, the pose change degree is calculated, and the key frame point cloud and LIDAR odometer are confirmed according to the pose change degree; And based on the key frame point cloud and the current target laser point cloud, a stitching map and factor information corresponding to the stitching map are obtained; Determine whether all laser point cloud data have been mapped; If yes, the factor information corresponding to the current spliced ​​map is output; If not, the LIDAR odometer is pre-integrated to obtain a new IMU odometer, and the new IMU odometer is used as the IMU odometer for the next mapping operation, and the next mapping operation is performed.

7. The multi-vehicle collaborative laser mapping method according to claim 6, characterized in that: The key frame point cloud includes: a first key frame point cloud and a second key frame point cloud; the LIDAR odometer includes: a first LIDAR odometer and a second LIDAR odometer; the step of confirming the key frame point cloud and the LIDAR odometer according to the degree of change of posture includes: The vehicle makes a judgment based on the degree of change of the posture; If the degree of change of the posture is greater than the degree of change threshold and there is a previous target laser point cloud to perform a mapping operation, the vehicle uses the current target laser point cloud as the first initial key frame point cloud and generates an odometer factor at the same time; the vehicle determines the closed loop factor based on closed loop detection; the vehicle obtains the first GPS factor by converting the GPS data corresponding to the target laser point cloud; the vehicle optimizes the first initial key frame point cloud based on the odometer factor, the closed loop factor and the first GPS factor to obtain the first key frame point cloud, and determines the first LIDAR odometer based on the first key frame point cloud; If the posture change is less than the change threshold, or there is no previous target laser point cloud to perform mapping operation, the vehicle uses the current target laser point cloud as the second initial key frame point cloud, converts the GPS data corresponding to the target laser point cloud, and obtains the second GPS factor; the vehicle optimizes the second initial key frame based on the second GPS factor, obtains the second key frame point cloud, and determines the second LIDAR odometer based on the second key frame point cloud.

8. A laser mapping device for multi-vehicle collaboration, characterized in that: include: Data acquisition module, map selection module, map deletion module and map splicing module; The data acquisition module is used to acquire first three-dimensional point cloud maps corresponding to a plurality of vehicles; wherein each first three-dimensional point cloud map corresponds to a vehicle; and each of the vehicles generates a first three-dimensional point cloud map in different areas of the area to be mapped; The map selection module is used to determine the map to be deleted from all the first three-dimensional point cloud maps; The map deletion module is used to delete the map to be deleted to obtain a second three-dimensional point cloud map; The map stitching module is used to perform map stitching based on the positioning data and posture data of each of the second three-dimensional point cloud maps to complete the construction of the three-dimensional point cloud map of the area to be mapped.

9. A computer terminal device, characterized in that: The method comprises a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, a laser mapping method for multi-vehicle collaboration as described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute a multi-vehicle collaborative laser mapping method as described in any one of claims 1 to 7.

Images