Two-dimensional code mapping method, device and equipment

Abstract

The invention discloses a two-dimensional code mapping method, which comprises the following steps of: continuously identifying two adjacent two-dimensional code nodes in a map building process, and acquiring poses corresponding to the two two-dimensional code nodes; constructing a first cost function and a second cost function through the poses corresponding to the two two-dimensional code nodes to obtain a first residual error and a second residual error respectively; and fusing the first residual error and the second residual error into the constructed map through a Ceres algorithm so as to optimize the constructed map. According to the main technical scheme and effects, after the two-dimensional code information is fused, the method uses the positioning data of the two-dimensional code as a constraint, the map generation precision is optimized, the effective integration of the two-dimensional code and the SLAM (Synchronous Localization and Mapping) point cloud map enables the map positioning to be more accurate, and especially in a complex or irregular environment, the map generation precision is improved. The two-dimensional code data can significantly reduce positioning errors, and the mapping efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the field of mapping, and in particular, to a method, device, and equipment for generating a map using two-dimensional codes. Background Art

[0002] With the rapid development of the economy and technology, the application demand for mobile robots in various industries is increasing continuously. Simultaneous Localization and Mapping (SLAM) technology, as the core challenge in the mobile robot technology system, mainly solves the problem of the robot's real-time positioning and map construction in an unknown environment.

[0003] Currently, the demand for domestic AGV (Automated Guided Vehicle) robots continues to grow, mainly concentrated in industries such as manufacturing and logistics, and the application of AGVs in these fields is becoming more and more extensive. However, when generating a map in a large-scale scenario and a mixed scenario (including information such as two-dimensional codes), there are still problems such as slow map generation speed and lack of effective integration of two-dimensional code information; in the existing large-scale scenario mapping technology, lidar (LiDAR) or cameras are usually used for environmental perception, combined with the SLAM (Simultaneous Localization and Mapping) algorithm for mapping and positioning. In an actual scenario, the robot may encounter an environment containing two-dimensional codes, and the two-dimensional code positioning can provide additional auxiliary information. The existing solutions fail to effectively utilize the two-dimensional code information in map generation, affecting the positioning accuracy and also resulting in the ineffective use of two-dimensional code information during positioning. Summary of the Invention

[0004] Object of the Invention: The object of the present invention is to solve the defects in the prior art and provide a method and device for generating a map using two-dimensional codes.

[0005] Technical Solution:

[0006] In a first aspect, the present application proposes a method for generating a map using two-dimensional codes, including the steps of:

[0007] Continuously identifying two adjacent two-dimensional code nodes during the process of map construction to obtain the poses corresponding to the two two-dimensional code nodes;

[0008] Constructing a first cost function and a second cost function through the poses corresponding to the two two-dimensional code nodes, and respectively obtaining a first residual and a second residual;

[0009] Using the Ceres algorithm to fuse the first residual and the second residual into the constructed map to optimize the constructed map.

[0010] Preferably, the poses corresponding to the two two-dimensional codes include the two-dimensional code coordinate pose and the radar coordinate pose.

[0011] Preferably, the first cost function includes:

[0012] Construct the transformation matrix between two QR code nodes as follows:

[0013] ;

[0014] where R is the rotation matrix and t is the translation vector;

[0015] Then the relative pose transformation between the nodes can be expressed as:

[0016] ;

[0017] where the poses of two QR code nodes A and B are respectively and , represents the transformation of node B relative to node A;

[0018] Calculate the first residual, and the first residual includes a translation residual, an angular residual, and a weighted residual.

[0019] Preferably, calculating the translation residual, the angular residual, and the weighted residual includes:

[0020] Obtain the position coordinates and rotation angles of QR code nodes A and B through the poses of two QR code nodes A and B;

[0021] Rotate through the position coordinates of node A to obtain the projection of node B in the coordinate system of node A, and the formula is as follows:

[0022] ;

[0023] where is the projection of node B in the coordinate system of node A, is the rotation angle of node A relative to the origin of the QR code coordinate system, is the position of node A on the coordinate system, is the rotation angle of node B relative to the origin of the QR code coordinate system, is the position of node B on the coordinate system, is the rotation matrix.

[0024] Preferably, calculating the translation residual, the angular residual, and the weighted residual further includes:

[0025] Calculate the translation residual, and the formula is as follows:

[0026]

[0027] where is the position difference between node A and node B on the map;

[0028] The calculated angular residual is as follows:

[0029] ;

[0030] where, is the angular value between node A and node B;

[0031] The calculated weighted residual is as follows:

[0032] ;

[0033] where, 𝑊 is the square root information matrix.

[0034] Preferably, the second cost function includes:

[0035] The poses of two QR code nodes A and B include the poses in the QR code coordinate system and the poses in the radar coordinate system;

[0036] The pose change in the QR code coordinate system is obtained from the poses of nodes A and B in the QR code coordinate system:

[0037] The pose change in the QR code coordinate system is obtained from the poses of nodes A and B in the radar coordinate system;

[0038] The calculated second residual is as follows:

[0039] ;

[0040] where, is the pose change in the QR code coordinate system obtained from the poses of nodes A and B in the QR code coordinate system and ; ;

[0041] is the pose change in the radar coordinate system obtained from the poses of nodes A and B in the radar coordinate system and and is obtained by unifying to the QR code coordinate system through ,

[0042] Preferably, the first residual and the second residual are fused into the constructed map by using the Ceres algorithm to optimize the constructed map, including:

[0043] Ceres uses a non - linear optimization algorithm to iteratively solve, minimizing the sum of the squares of the first residual and the second residual in the first cost function and the second cost function, as follows:

[0044] ;

[0045] ​Among them, residual1(i) and residual2(i) are the residuals under the QR code map and the lidar map respectively, and i represents different node pairs.

[0046] Preferably, it further includes: performing voxel filtering on the optimized map, including:

[0047] Dividing the constructed map into regular voxel cube units;

[0048] Calculating the centroid for each voxel, and using the centroid point to replace all the points within the voxel for calculation.

[0049] In a second aspect, an embodiment of the present invention provides a QR code rapid mapping device, including the method described in the above embodiment, including:

[0050] An acquisition unit, configured to continuously identify two adjacent QR code nodes during the process of constructing a map, and acquire the poses corresponding to the two QR code nodes;

[0051] A processing unit, configured to construct a first cost function and a second cost function through the poses corresponding to the two QR code nodes, and respectively obtain a first residual and a second residual;

[0052] An optimization unit, configured to use the Ceres algorithm to fuse the first residual and the second residual into the constructed map to optimize the constructed map.

[0053] In a third aspect, an embodiment of the present invention provides an electronic device, including a processor and a memory. Among them, the memory is used to store one or more computer programs; when the one or more computer programs stored in the memory are executed by the processor, the electronic device can implement the method of any possible design in the first aspect above.

[0054] In a fourth aspect, the present invention provides a computer-readable storage medium, in which a computer program is stored, and when the computer program is executed by a processor, the method in any one of the above embodiments is implemented.

[0055] In a fifth aspect, an embodiment of the present invention further provides a computer program product, when the computer program product runs on an electronic device, the electronic device is enabled to execute the method of any possible design in any one of the above aspects.

[0056] Beneficial effects: After fusing the QR code information, the method uses the positioning data of the QR code as a constraint to optimize the accuracy of map generation. The effective integration of the QR code and the SLAM (Simultaneous Localization and Mapping) point cloud map makes the positioning of the map more accurate. Especially in complex or irregular environments, the QR code data can significantly reduce the positioning error and improve the mapping efficiency;

[0057] Through the optimized map optimization algorithm and QR code information fusion, the stability of the mapping process is improved, the cumulative error is reduced, and it is ensured that high-quality mapping effects can be maintained even in large-scale environments;

[0058] The use of voxel filtering technology simplifies the processing of point cloud data, significantly reduces the data storage space requirements, and at the same time improves the storage and processing efficiency. Especially when dealing with complex large scenes, the system burden is reduced. By optimizing the point cloud data through voxel filtering, the number of stored point clouds is greatly reduced, while ensuring the map accuracy, a large amount of memory and computing resources are saved, and the system processing speed is improved. Description of the Drawings

[0059] Figure 1 It is a schematic diagram of the method framework provided by the present invention;

[0060] Figure 2 It is a schematic diagram before and after voxel filtering of the present invention;

[0061] Figure 3 It is a schematic diagram of the effects before and after voxel filtering of the present invention;

[0062] Figure 4 It is a schematic diagram of a device provided by the present invention;

[0063] Figure 5 It is a schematic diagram of a device provided by the present invention. Detailed Embodiments

[0064] To make the technical solutions of the present invention clearer, the following further describes the present invention in detail with specific embodiments in conjunction with the accompanying drawings.

[0065] Embodiment 1

[0066] To make the purpose, technical solutions and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention. Unless otherwise defined, the technical terms or scientific terms used herein shall have the ordinary meaning understood by those of ordinary skill in the art in the field to which the present invention belongs. The words such as "including" used herein mean that the elements or objects appearing before the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects.

[0067] Aiming at the problems existing in the prior art, as Figures 1-3 shown, the present invention provides a QR code mapping method, including the steps:

[0068] S101. During the process of constructing the map, continuously identify two adjacent QR code nodes, and obtain the poses corresponding to the two QR code nodes. During the map construction process, the robot AGV continuously scans the QR codes in the environment to identify two adjacent QR code nodes. For each pair of adjacent QR code nodes, the robot obtains their corresponding pose information, that is, the relative position and orientation of the QR code nodes.

[0069] S102. Construct the first cost function and the second cost function through the poses corresponding to the two QR code nodes, and respectively obtain the first residual and the second residual. For each pair of adjacent QR code nodes, based on their pose information, construct the first cost function and the second cost function. These two cost functions will respectively measure the error related to the position of the QR code nodes in the trajectory optimization of the robot.

[0070] The first cost function: used to calculate the translation error or pose deviation, reflecting the difference in the robot's position estimation.

[0071] The second cost function: used to measure the rotation error or angle change, and further optimize the relative orientation between the robot and the QR code nodes.

[0072] Through the calculation of the cost function, two residuals (the first residual and the second residual) are obtained. These two residuals describe the error between the current position of the robot and the QR code nodes.

[0073] S103. Use the Ceres algorithm to fuse the first residual and the second residual into the constructed map to optimize the constructed map. The first residual and the second residual are fused into the map optimization through the Ceres algorithm. The Ceres algorithm is an optimization method commonly used for nonlinear least squares problems, which can optimize the accuracy of the map by minimizing the residuals.

[0074] The Ceres algorithm adjusts the pose of the robot in the map in an iterative manner to minimize the position and orientation errors. The optimized map will be more accurate.

[0075] Specifically, during the process of positioning and mapping, multiple QR codes will be set on the factory floor. When the AGV cart passes by, it will scan the information of the QR codes. In map optimization, the robot needs to use multiple sensors (such as lidar) for positioning and mapping, and the QR code pose data provides additional positioning constraints. These QR codes are usually embedded in the scene. The robot identifies and locates the QR codes, thereby providing additional information for map optimization and reducing the influence of errors of other sensors.

[0076] For each pair of adjacent QR code nodes, the system calculates the relative pose between them, i.e., the offset of the robot relative to the center of the QR code. This relative pose information can help determine the position and orientation changes between two QR codes, serving as a constraint in the optimization graph. This method enables the effective fusion of QR code information with other sensor data (such as lidar), further improving the map accuracy.

[0077] By constructing two cost functions, the optimization algorithm calculates the residuals of the QR code pose constraints and optimizes the robot's trajectory and position by minimizing these residuals. This process can significantly improve the accuracy of graph optimization, and the optimized pose enhances the accuracy and consistency of the global map.

[0078] Since the pose data provided by the QR code can serve as a constraint, when the optimization algorithm solves the robot's trajectory, it can better adjust the pose, reducing the position deviation caused by environmental complexity or sensor errors. The pose data of the QR code, especially in an irregular environment, can help improve the system's ability to estimate the robot's trajectory, thereby enhancing the overall mapping accuracy.

[0079] In some preferred embodiments, the poses corresponding to the two QR codes include the QR code coordinate pose and the radar coordinate pose. The pose of the QR code in the QR code coordinate system is usually the position and orientation relative to the origin of the QR code (such as the center of the QR code), which is usually obtained by the robot through the QR code recognition algorithm and can be used to provide constraints on the position and direction.

[0080] The radar coordinate pose is the pose obtained by the robot through lidar or other sensors, representing the position and direction of the robot in the radar coordinate system. The radar coordinate system is a reference frame in the space sensed by the robot through sensors such as lidar.

[0081] In some preferred embodiments, the first cost function includes:

[0082] Construct the transformation matrix between two QR code nodes as follows:

[0083] ;

[0084] where R is the rotation matrix and t is the translation vector;

[0085] Then the relative pose transformation between the nodes can be expressed as:

[0086] ;

[0087] where the poses of two QR code nodes A and B are respectively and , Represents the transformation of node B relative to node A, where node A and node B are adjacent nodes;

[0088] The first residual is calculated, and the first residual includes a translation residual, an angle residual, and a weighted residual.

[0089] In some preferred embodiments, calculating the translation residual, the angle residual, and the weighted residual includes:

[0090] Obtain the position coordinates and rotation angles of the two QR code nodes A and B through the poses of the two QR code nodes A and B;

[0091] Rotate through the position coordinates of node A to obtain the projection of node B in the coordinate system of node A. The formula is as follows:

[0092] ;

[0093] Where, is the projection of node B in the coordinate system of node A, is the rotation angle of node A relative to the origin of the QR code coordinate system, is the position of node A on the coordinate system, is the rotation angle of node B relative to the origin of the QR code coordinate system, is the position of node B on the coordinate system, is the rotation matrix. By optimizing the conversion between coordinate systems through the rotation matrix, the transformation of the QR code position and the position information between nodes are fused, thereby reducing errors in map construction and path planning. By calculating the translation error and the rotation error, the accuracy of the map can be optimized.

[0094] In some preferred embodiments, calculating the translation residual, the angle residual, and the weighted residual further includes:

[0095] Calculate the translation residual. The formula is as follows:

[0096]

[0097] Where, is the position difference between node A and node B on the map, which is a predefined prior position difference and is the relative translation transformation of the pose solved by the QR code. This formula calculates the translation error of node B relative to node A;

[0098] Calculate the angle residual. The formula is as follows:

[0099] ;

[0100] Where, is the angular value between node A and node B, the predefined angular difference, the relative transformation of pose rotation calculated by decoding the QR code. This formula calculates the angular difference between node A and node B. Since the angle is periodic, it is necessary to standardize the angular difference to keep it within the range of [−π, π];

[0101] The weighted residual is calculated, and the formula is as follows:

[0102] ;

[0103] where 𝑊 is the square root information weighting matrix, which is used to weight the translation residual and the angular residual according to importance or confidence. Calculate the error between robots or nodes, especially when converting between different coordinate systems, calculate the relative translation and rotation errors between nodes and weight these errors; during map construction or path optimization, accurately calculate the errors and adjust different errors through the weighting matrix to achieve higher-precision positioning and mapping.

[0104] In some preferred embodiments, the second cost function includes:

[0105] The poses of two QR code nodes A and B include the poses in the QR code coordinates and the poses in the radar coordinates;

[0106] Obtain the pose change in the QR code coordinate system through the poses of nodes A and B in the QR code coordinates:

[0107] Obtain the pose change in the QR code coordinate system through the poses of nodes A and B in the radar coordinates;

[0108] The second residual is calculated, and the formula is as follows:

[0109] ;

[0110] where is the pose change in the QR code coordinate system obtained through the poses of nodes A and B in the QR code coordinates and ; ;

[0111] is the pose change in the radar coordinate system obtained through the poses of nodes A and B in the radar coordinate system and ; , and is obtained by unifying to the QR code coordinate system through .

[0112] In some preferred embodiments, the first residual and the second residual are fused into the constructed map by the Ceres algorithm to optimize the constructed map, including:

[0113] Ceres uses a non - linear optimization algorithm to iteratively solve, minimizing the sum of the squares of the first residual and the second residual in the first cost function and the second cost function. The formula is as follows:

[0114] ;

[0115] where residual1(i) and residual2(i) are the residuals under the QR code map and the lidar map respectively, and i represents different node pairs.

[0116] In some preferred embodiments, such as Figures 2-4 , it further includes: performing voxel filtering on the optimized map, including:

[0117] Dividing the constructed map into regular voxel cube units;

[0118] Calculating the centroid for each voxel, and using the centroid point to replace all the points within the voxel for calculation;

[0119] In large - scale SLAM mapping or when repeatedly scanning the same scene, the quantity of point cloud data usually increases rapidly. Especially in complex environments, as the scanning range expands or the same scene is scanned multiple times, the volume of the point cloud map gradually becomes huge, thus occupying a large amount of memory and computing resources. This not only increases the storage overhead but also significantly reduces the computing efficiency, affecting the speed and quality of subsequent processing;

[0120] To solve this problem, the point cloud data is simplified through voxel filtering (Voxel Grid Filtering), thereby reducing the data volume while retaining the global shape and structural features of the point cloud;

[0121] The three - dimensional space is divided into regular voxel grids. Each voxel is a cube unit of the same size, and the size of the voxel is set by the user. Smaller voxels can retain more detailed information, while larger voxels can significantly reduce the data volume but may sacrifice some details;

[0122] By dividing with voxel grids with a side length of 2 cm, all the points within each voxel (the voxel is each of the small grids divided in the previous text) will be simplified. A common simplification method is to calculate the centroid of all the points within the voxel and use this centroid to represent all the points within the voxel;

[0123] such as Figure 2 and Figure 3As shown, (a) shows that before voxel filtering, the space has been divided into grids, and (b) shows that after voxel filtering, there is only one point in each grid. This simplified processing can not only effectively reduce the number of point clouds, but also smooth the noise to a certain extent, thereby improving the quality of point cloud data;

[0124] In some examples, the number of point clouds before filtering is 56,485, and the number of point clouds after filtering is 21,880. After voxel filtering, the number of point clouds is reduced by about 60%, greatly reducing the space occupancy, and the mapping accuracy is not affected.

[0125] In some preferred embodiments, the loop closure detection process is accelerated through parallel computing, and multiple threads are used to execute the loop closure detection task simultaneously, thereby improving the loop closure detection efficiency; during the large-scale scene mapping process, the front end real-time calls and processes a large amount of point cloud data and draws it into a map; by temporarily blocking the front-end and back-end data transmission channels, the delay of data interaction during the mapping process is reduced, thereby accelerating the mapping process; after the mapping parameters are set, the system temporarily blocks the front-end and back-end data transmission to avoid the influence of data transmission delay on the mapping real-time performance;

[0126] Among them, through multi-threaded parallel computing, loop closure detection can effectively meet the real-time mapping requirements in large-scale and complex environments;

[0127] Among them, the mechanism of temporarily blocking the front-end and back-end data transmission channels can significantly improve the speed and real-time performance of large-scale scene map construction, and ensure the stable operation of the system in complex environments;

[0128] Among them, through setting reasonable isolation strategies for the data transmission between the front end and the back end, unnecessary data exchange is reduced, and the overall mapping efficiency is improved.

[0129] In a second aspect, an embodiment of the present invention provides a two-dimensional code fast mapping device, combined with Figure 4 , including the method described in the above embodiment, including:

[0130] An acquisition unit 301, configured to continuously identify two adjacent two-dimensional code nodes during the process of constructing a map, and acquire the poses corresponding to the two two-dimensional code nodes;

[0131] A processing unit 302, configured to construct a first cost function and a second cost function through the poses corresponding to the two two-dimensional code nodes, and respectively obtain a first residual and a second residual;

[0132] An optimization unit 303, configured to fuse the first residual and the second residual into the constructed map through the Ceres algorithm to optimize the constructed map.

[0133] All relevant content of each step involved in the above method embodiments can be cited in the function description of the corresponding functional module, and will not be elaborated here.

[0134] In some other embodiments of the present invention, embodiments of the present invention disclose an electronic device, such as Figure 5 shown, the electronic device may include: one or more processors 401; a memory 402; a display 403; one or more applications (not shown); and one or more computer programs 404. The above components may be connected through one or more communication buses 405. Wherein the one or more computer programs 404 are stored in the memory 402 and configured to be executed by the one or more processors 401. The one or more computer programs 404 include instructions, and the above instructions may be used to execute the steps in Figures 1 to 2 and the corresponding embodiments.

[0135] Through the description of the above embodiments, those skilled in the art can clearly understand that for the convenience and conciseness of description, only the above division of each functional module is used as an example. In actual applications, the above functions can be allocated to different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. The specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated here.

[0136] In each embodiment of the present invention, each functional unit can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

[0137] If the above integrated unit 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 this understanding, the technical solution of the embodiments of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor to execute all or part of the steps of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: various media such as flash memory, mobile hard disk, read-only memory, random access memory, magnetic disk, or optical disc that can store program codes.

[0138] As described above, it is only the specific implementation manner of the embodiments of the present invention. However, the protection scope of the embodiments of the present invention is not limited thereto. Any changes or substitutions within the technical scope disclosed in the embodiments of the present invention should be covered within the protection scope of the embodiments of the present invention. Therefore, the protection scope of the embodiments of the present invention shall be subject to the protection scope of the claims.

Claims

1. A method for generating a map using two-dimensional codes, characterized in that, Including the steps: During the process of constructing the map, continuously identify two adjacent QR code nodes and obtain the poses corresponding to the two QR code nodes; Construct the first cost function and the second cost function through the poses corresponding to the two QR code nodes, and respectively obtain the first residual and the second residual; Use the Ceres algorithm to fuse the first residual and the second residual into the constructed map to optimize the constructed map.

2. The method according to claim 1, wherein The poses corresponding to the two QR codes include the QR code coordinate pose and the radar coordinate pose.

3. The method according to claim 2, wherein The first cost function includes: Construct the transformation matrix between the two QR code nodes as follows: ； where, R is the rotation matrix and t is the translation vector; Then the relative pose transformation between the nodes can be expressed as: ； wherein, the poses of two QR code nodes A and B are respectively and , represents the transformation matrix of node B relative to node A; Calculate the first residual, and the first residual includes the translation residual, the angle residual and the weighted residual.

4. The method according to claim 3, characterized in that Calculating the translation residual, the angle residual and the weighted residual includes: Obtain the position coordinates and the rotation angle of the QR code nodes A and B through the poses of the two QR code nodes A and B; Rotate through the position coordinates of node A to obtain the projection of node B in the coordinate system of node A. The formula is as follows: ； Among them, is the projection of node B in the coordinate system of node A, is the rotation angle of node A relative to the origin of the QR code coordinate system, is the position of node A on the coordinate system, is the rotation angle of node B relative to the origin of the QR code coordinate system, is the position of node B on the coordinate system, is the rotation matrix.

5. The method according to claim 4, wherein Calculating the translation residual, the angle residual and the weighted residual also includes: Calculate the translation residual, and the formula is as follows: ; wherein, is the position difference between node A and node B on the map; Calculate the angle residual, and the formula is as follows: ； Among them, is the angular value between node A and node B; Calculate the weighted residual, and the formula is as follows: ； where, 𝑊 is the square root information matrix.

6. The method according to claim 4, wherein The second cost function includes: The poses of the two QR code nodes A and B include the poses in the QR code coordinates and the poses in the radar coordinates; Obtain the pose change in the QR code coordinate system through the poses of nodes A and B in the QR code coordinates; Obtain the pose change in the QR code coordinate system through the poses of nodes A and B in the radar coordinates; Calculate the second residual, and the formula is as follows: ； Among them, is the pose of nodes A and B under the QR code coordinates and obtains the pose change in the QR code coordinate system; The poses of nodes A and B in the radar coordinate system and to obtain the pose change in the radar coordinate system , and through Unify to the QR code coordinate system to obtain.

7. The method according to claim 1, wherein Using the Ceres algorithm to fuse the first residual and the second residual into the constructed map to optimize the constructed map includes: Ceres uses a non-linear optimization algorithm to iteratively solve, minimizing the sum of the squares of the first residual and the second residual in the first cost function and the second cost function. The formula is as follows: ； where, residual1(i) and residual2(i) are the residuals under the QR code map and the lidar map respectively, and i represents different node pairs.

8. The method according to claim 1, wherein It also includes: Perform voxel filtering on the optimized map, including: Divide the constructed map into regular voxel cube units; Calculate the centroid for each voxel, and use the centroid point to replace all the points within the voxel for calculation.

9. A QR code rapid mapping device, comprising the method according to any one of claims 1-8, characterized in that, Including: An acquisition unit for continuously identifying two adjacent QR code nodes during the process of constructing the map and obtaining the poses corresponding to the two QR code nodes; A processing unit for constructing the first cost function and the second cost function through the poses corresponding to the two QR code nodes and respectively obtaining the first residual and the second residual; An optimization unit for using the Ceres algorithm to fuse the first residual and the second residual into the constructed map to optimize the constructed map.

10. An electronic device, characterized in that, Comprising a memory and a processor, a computer program capable of running on the processor is stored on the memory. When the computer program is executed by the processor, the processor is caused to implement the method according to any one of claims 1 to 8.

Images