Method, device, system and medium for edge positioning of road rollers based on lidar

Abstract

This application relates to a method, device, system, and medium for edge positioning of a road roller based on lidar, which achieves high positioning accuracy during edge positioning operations. The lidar-based method for edge positioning of a road roller includes: acquiring point clouds obtained by multi-line lidar scanning of the road surface and curb; dividing the point cloud into multiple first grid cells along the horizontal direction, and acquiring the center point cloud and point cloud direction of each first grid cell; clustering the multiple first grid cells according to the center point cloud and point cloud direction of each first grid cell to obtain a curb region and a road surface region; acquiring a curb point cloud set and a road surface point cloud set according to the curb region and the road surface region; performing linear fitting on the curb point cloud set and the road surface point cloud set to obtain a curb fitting line and a road surface fitting line, and selecting the point cloud closest to the intersection point and within the curb point cloud set as the boundary point between the road surface and the curb; and calculating the edge positioning distance and pose tilt angle based on the boundary point.

Description

Technical Field

[0001] This application belongs to the field of road roller technology, and in particular relates to a method, device, system and medium for edge positioning of road rollers based on lidar. Background Technology

[0002] Asphalt pavement laying in road engineering requires the use of double-drum or single-drum rollers to compact the asphalt layer. During the compaction process, it is necessary to ensure the accuracy of the distance between the roller and the curbstone. This avoids problems such as excessive under-compaction, chipping, and damaged edges due to excessive distance, and damage to the curbstone due to insufficient distance.

[0003] In existing technologies, for manned road rollers, the operator can determine the edge-fitting distance by observing the distance between the roller and the curb. However, this method has low automation, cannot guarantee the positioning accuracy of the edge-fitting distance, and cannot be applied to unmanned road rollers. For unmanned road rollers, the roller's travel path and road edge can be identified by placing markers on the road. However, this method requires a large number of markers, making engineering implementation difficult, and positioning accuracy cannot be guaranteed when operating at close range. Alternatively, non-contact sensors can be used to measure the distance between the roller and the curb to obtain the edge-fitting distance and tilt angle. However, this method has poor adaptability and is only suitable for flat road surfaces and right-angled curbs. If the road surface is uneven, the non-contact sensors will be blocked when measuring the distance. Or, if the curb is sloping, the non-contact sensors cannot accurately measure the edge-fitting distance between the roller and the curb. Summary of the Invention

[0004] In view of this, the embodiments of this application provide a method, device, system and medium for edge positioning of road rollers based on lidar, so as to solve the problem of low positioning accuracy when positioning road rollers at the edge in the prior art.

[0005] The first aspect of this application provides a method for positioning a road roller along a curb based on lidar, comprising: acquiring point clouds obtained by multi-line lidar scanning of the road surface and curb; dividing the point clouds into multiple first grid cells in a horizontal direction, acquiring the center point cloud and point cloud direction of the first grid cells, wherein the horizontal direction is the direction from the road surface to the curb, and the coordinates of the center point cloud of the first grid cells are the average coordinates of the point clouds within the first grid cells; clustering the multiple first grid cells according to the center point cloud and point cloud direction of the first grid cells to obtain a curb region and a road surface region; acquiring a curb point cloud set and a road surface point cloud set according to the curb region and the road surface region; performing linear fitting on the curb point cloud set and the road surface point cloud set to obtain a curb fitting line and a road surface fitting line, and selecting the point cloud closest to the intersection point and in the curb point cloud set as the boundary point between the road surface and the curb based on the boundary point; and calculating the edge-adhering distance and pose tilt angle of the road roller according to the boundary point.

[0006] In one embodiment, the step of clustering multiple first grid cells to obtain a curb area and a road surface area based on the center point cloud and point cloud direction of the first grid cell includes: clustering the first grid cell into a road surface area when the height difference between the center point coordinates of the first grid cell and the installation height of the multi-line lidar is less than a preset threshold, the height difference between the center point cloud of the first grid cell and the center point cloud of the previous first grid cell is less than a preset height difference, and the angle difference between the point cloud direction of the first grid cell and the point cloud direction of the previous first grid cell is less than a preset angle difference; and clustering the first grid cell into a curb area when the height difference between the center point cloud of the first grid cell and the center point cloud of the previous first grid cell is greater than a preset height difference, and the angle difference between the point cloud direction of the first grid cell and the point cloud direction of the previous first grid cell is greater than a preset angle difference.

[0007] In one embodiment, obtaining the curb point cloud and the road point cloud based on the curb area and the road surface area includes: dividing the curb area into multiple second grid units in a horizontal direction, obtaining the center point cloud and point cloud direction of the second grid units, wherein the second grid units are smaller than the first grid units; and clustering the multiple second grid units based on the center point cloud and point cloud direction of the second grid units to obtain the curb point cloud and the road surface point cloud.

[0008] In one embodiment, obtaining the center point cloud and point cloud orientation of the first grid cell includes: obtaining the point cloud orientation of the first grid cell using principal component analysis.

[0009] In one embodiment, after acquiring the point cloud obtained by multi-line lidar scanning of the road surface and curbstone, the method further includes: filtering the point cloud, wherein the filtered point cloud p is:

[0010]

[0011] Where n is the line number of the scanning line of the multi-line lidar, i is the point number of the scanning line, H is the ground clearance of the multi-line lidar, th1 and th2 are the preset height thresholds, x, y, and z are the coordinates of the initial point cloud in the O-XYZ coordinate system, the origin of the O coordinate system is the optical center of the multi-line lidar, the Z-axis is vertical to the road surface and downwards is positive, the Y-axis is parallel to the direction of travel of the road roller, and the X-axis is parallel to the wheel axle centerline of the road roller.

[0012] In one embodiment, the step of calculating the edge-fitting distance and tilt angle of the road roller based on the boundary point includes: performing planar fitting on the point cloud of the curb stone to obtain the normal vector of the plane where the curb stone is located; and calculating the tilt angle of the road roller based on the normal vector of the plane where the curb stone is located.

[0013] A second aspect of this application provides a road roller edge positioning device based on lidar, comprising: a scanning module for acquiring point clouds obtained by multi-line lidar scanning of the road surface and curb; a division module for dividing the point cloud into multiple first grid units in a horizontal direction, acquiring the center point cloud and point cloud direction of the first grid unit, wherein the horizontal direction is the direction from the road surface to the curb, and the coordinates of the center point cloud of the first grid unit are the average coordinates of the point cloud within the first grid unit; and a clustering module for clustering based on the center point cloud and point cloud direction of the first grid unit. The system performs clustering on multiple first grid cells to obtain a curb region and a road surface region; an acquisition module is used to acquire a curb point cloud and a road surface point cloud based on the curb region and the road surface region; a boundary module is used to perform straight line fitting on the curb region and the road surface region to obtain a curb fitting line and a road surface fitting line, and selects the point cloud closest to the intersection point and in the curb point cloud as the boundary point between the road surface and the curb stone based on the intersection point; a calculation module is used to calculate the edge-fitting distance and pose tilt angle of the road roller based on the boundary point.

[0014] A third aspect of this application provides a road roller edge positioning system based on lidar, including the multi-line lidar and the road roller edge positioning device as described in the second aspect, wherein the ground clearance H of the multi-line lidar and the side distance W between the multi-line lidar and the side of the road roller satisfy the following constraint relationship:

[0015]

[0016] Where k is the scanning angle resolution of the multi-line lidar.

[0017] In one embodiment, the system further includes an adjustable bracket connected to the multi-line lidar for adjusting the ground clearance H and the lateral distance W of the multi-line lidar.

[0018] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method as described in any of the first aspects.

[0019] This application discloses a method, device, system, and medium for positioning a road roller along its edge based on lidar. Compared with existing technologies, the advantages of this method are as follows:

[0020] (1) In this embodiment, a multi-line lidar is used to scan the road surface and curbstone. The scanned point cloud is divided into grids, and the first grid cells are clustered to roughly locate the curbstone and road surface areas. Linear fitting and intersection processing are performed on the point cloud sets of the curbstone and road surface areas. The point closest to the intersection point in the curbstone point cloud set is determined as the boundary point between the road surface and the curbstone, thereby accurately dividing the curbstone and the road surface. When there is a stone protrusion on the road surface close to the curbstone, the traditional method of traversing the point cloud may select the boundary point in front of the stone, resulting in a large calculated edge distance. The method of this embodiment can accurately obtain the boundary point, and thus accurately obtain the edge distance and pose angle between the road roller and the curbstone.

[0021] (2) In this embodiment, the scanned point cloud is divided into grids. The first grid unit is clustered based on its center point cloud and point cloud direction, which enables rapid coarse localization of the road surface and curbstone. Furthermore, the coarsely localized curbstone area is divided into grids, and the second grid unit is clustered based on its center point cloud and point cloud direction, further refining the localization of the curbstone and road surface areas. Compared to the traditional method of traversing the scanned point cloud, this method significantly saves processing time and improves processing efficiency. Moreover, grid division can ignore the undulations of single points within the first grid unit, avoiding interference with the processing results due to protruding stones.

[0022] (3) In this embodiment, point clouds with a distance less than or equal to a preset distance threshold from the fitted line of the curb area are obtained from all point clouds to form a curb point cloud set, thereby obtaining a more complete curb. For sloping roads, if the boundary point between the road surface and the curb is divided at the center of the first grid cell, the boundary point may be submerged by the road surface and thus clustered on the ground. In this embodiment, a reverse search can be performed to obtain the complete curb area, and the more point clouds of the curb, the higher the accuracy of the obtained edge distance and pose tilt angle.

[0023] (4) The embodiments of this application filter the point cloud obtained by scanning, filter out points that are blown away due to water accumulation or other factors, eliminate interference caused by point clouds that are seriously deviated from the road surface and curbstone, and reduce the amount of data calculation and improve the calculation speed.

[0024] (5) The embodiments of this application use a planar fitting method for the point cloud data of the curbstone, which can be adapted to the calculation of the pose and tilt angle when the curbstone is a right angle or a slope.

[0025] (6) In the embodiments of this application, the ground height H and side distance W of the multi-line lidar meet the constraint conditions, which can ensure that the positioning accuracy of the edge distance is controlled within 1cm, avoid problems such as underpressure, slag falling and broken edge during edge operation, and meet the engineering implementation requirements of the road roller. Attached Figure Description

[0026] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0027] Figure 1 This is a schematic diagram of a road roller provided in an embodiment of this application;

[0028] Figure 2 This is a flowchart illustrating a method based on lidar provided in an embodiment of this application;

[0029] Figure 3 This is a schematic diagram of a multi-line lidar provided in an embodiment of this application;

[0030] Figure 4 This is a schematic diagram of a multi-line lidar scanning of the road surface and curbstone provided in an embodiment of this application;

[0031] Figure 5 This is a schematic diagram of the filtering process provided in the embodiments of this application;

[0032] Figure 6This is a schematic diagram of the center point cloud and point cloud direction of the first grid cell provided in an embodiment of this application;

[0033] Figure 7 This is a schematic diagram showing the intersection of the road surface fitting line and the roadside fitting line provided in the embodiments of this application;

[0034] Figure 8 This is a schematic diagram of the selection of the boundary point provided in the embodiments of this application;

[0035] Figure 9 This is a schematic diagram of the edge contact distance and pose tilt angle provided in the embodiments of this application;

[0036] Figure 10 This is a schematic diagram of planar fitting calculation of pose tilt angle provided in an embodiment of this application;

[0037] Figure 11 This is a schematic diagram of the height H above the ground and the side distance W of the multi-line lidar provided in the embodiments of this application;

[0038] Figure 12 This is a schematic diagram showing the constraint relationship between the ground clearance H and the side distance W provided in the embodiments of this application. Detailed Implementation

[0039] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0040] To illustrate the technical solution described in this application, specific embodiments are provided below.

[0041] like Figure 1 The diagram shows the structure of a road roller. During manned operation, the steel wheel structure creates blind spots on both sides, making it difficult for the operator to accurately observe the edge compaction distance. This can lead to under-compaction, crumbling, and damaged edges when the distance is too far; conversely, too close a distance can result in the wheel damaging the curb. In unmanned operation, since edge compaction is entirely controlled by the roller itself, the lack of accurate edge compaction distance and the positional angle of the wheels relative to the curb can cause the roller to malfunction or result in similar issues as manned operation, such as under-compaction, crumbling, damaged edges, and curb damage.

[0042] To address the aforementioned problems in edge compaction operations using road rollers, the commonly used technical solutions are as follows:

[0043] (1) Positioning Vernier Method and Panoramic Image Method. The positioning vernier method uses an adjustable positioning vernier along the width of the steel wheel to indicate the position of the curbstone to the driver. The panoramic image method uses image acquisition devices around the roller to provide the driver with images of the roller and the surrounding area of ​​the curbstone. Both of these methods fall under the category of auxiliary positioning. The driver can observe and determine the position of the curbstone without using a probe, which can greatly reduce the difficulty of the driver's operation for edge positioning. However, the degree of automation is low and it cannot be applied to unmanned rollers.

[0044] (2) Path Edge Marking Method. This method uses multiple brightly colored marker poles to mark the road path and edges, allowing the vision system on the unmanned road roller to quickly identify the roller's travel path and road edges, thus achieving automated operation of the unmanned road roller. However, this method requires the deployment of a large number of marker poles and marking lines along the route, making engineering implementation difficult. Furthermore, it cannot guarantee positioning accuracy when operating close to the edge, and problems such as under-compaction, slag shedding, and damaged edges still exist.

[0045] (3) Multi-sensor method. This method uses at least eight non-contact sensors to directly measure the distance between the left and right sides of the front and rear compaction wheels and the curbstone, and obtains the edge-fitting distance and tilt angle based on geometric relationships. This method can achieve edge-fitting compaction operation for both manned and unmanned road rollers when the road surface to be compacted is relatively flat and the curbstone is right-angled. However, if encountering potholes or raised road surfaces, there will be problems with the inability to measure the distance between the compaction wheel and the curbstone due to obstruction. If the curbstone is sloping, there will be problems with inaccurate distance acquisition.

[0046] To address the problems existing in the prior art, this application provides a method, device, system, and medium for positioning a road roller along the edge using lidar. It employs a multi-line lidar to scan the road surface and curbstone, performs grid division and clustering on the obtained point cloud to roughly locate the curbstone and road surface areas. Furthermore, it performs linear fitting and intersection analysis on the point cloud sets of the curbstone and road surface areas, searching for the point cloud set closest to the intersection point within the curbstone area as the boundary point. This boundary point accurately locates the curbstone and road surface, thereby enabling precise calculation of the road roller's edge-adhering distance and tilt angle with the curbstone, resulting in high positioning accuracy for edge-adhering compaction operations.

[0047] like Figure 2 As shown, the first aspect of this application provides a method for positioning a road roller along its edge based on lidar, including the following steps:

[0048] S101. Obtain the point cloud obtained from multi-line lidar scanning of the road surface and curb stones.

[0049] This application embodiment can install one or more multi-line lidar sensors on the wheels of a road roller, such as... Figure 3 As shown, the multi-line lidar in this embodiment can specifically be a 4-line lidar. The 4-line lidar includes four scanning lines, including a 0° scanning line. The scanning angle resolution of each of the four scanning lines is 0.2°, meaning that the angle between two adjacent scanning points on the same scanning line and the origin O is 0.2°. There will be 5 scanning points within 1°. The plane containing the 0° scanning line coincides with the plane OXZ, while the other scanning lines form angles with the plane OXZ. The angles formed by the four scanning lines are 9°, 3°, 0°, and -0.3°, respectively. The resulting scanning surfaces are, in order, a 9° cone, a 3° cone, a 0° circular surface, and a -0.3° cone. Figure 4 As shown, in the projection formed on the ground, the 0° scan line is a straight line, and the other scan lines are arcs.

[0050] When a 4-line lidar is installed on the wheel of a road roller, the plane containing the 0° scanning line is perpendicular to the direction of travel of the road roller. In the O-XYZ coordinate system established by the lidar, the optical center of the lidar is the origin O of the coordinate system, the Z-axis is vertical and downward to the ground as positive, the Y-axis is parallel to the direction of travel of the road roller, and the X-axis is parallel to the axis of the road roller wheel.

[0051] When using a multi-line lidar to scan the road surface and curb, the scanning range of the multi-line lidar is set to x∈[-x1, x2], where x is the measurement range on the left and right sides centered on the multi-line lidar, and only the point cloud data within this range is detected; and according to the installation height H of the lidar, only the point cloud within the range of H-th1≤z≤H+th2 from the ground is considered, and the point cloud within this range includes the road and curb.

[0052] This application embodiment uses a multi-line lidar to scan the road surface and curb stones. Compared with the prior art that uses non-contact sensors to directly obtain distance, it can overcome the problem of being unable to measure when encountering potholes and raised road surfaces due to obstruction, and it can also overcome the problem of inaccurate distance acquisition when encountering sloping curb stones. It has good adaptability and high positioning accuracy. Moreover, it is easy to operate and highly automated. It can provide operation guidance for manned road rollers and is also conducive to the integrated use of unmanned road rollers.

[0053] After acquiring the point clouds of the curbstone and road surface, the point clouds are filtered to focus only on the point clouds within the range of H-th1≤z≤H+th2 from the ground, retaining the point clouds within the road and curbstone range to reduce subsequent computation and improve computation speed. Furthermore, it can filter out point clouds whose coordinates are seriously deviated from the road surface and curbstone due to water accumulation or other factors. For example, when there is water accumulation on the ground, the coordinates of some point clouds formed by multi-line lidar scanning of the ground may be (0,0,0), causing some point clouds to deviate from the road surface outline. In order to ensure the accuracy of subsequent calculations, it is necessary to filter the point clouds first.

[0054] In one implementation, after acquiring the point cloud obtained from multi-line lidar scanning of the road surface and curb stones, the method further includes:

[0055] The point cloud is filtered, and the filtered point cloud p is:

[0056]

[0057] Where n is the line number of the scanning line of the multi-line lidar, i is the point number of the scanning line, H is the ground clearance of the multi-line lidar, th1 and th2 are preset height thresholds, which can be the same or different, x, y, and z are the coordinates of the point cloud in the O-XYZ coordinate system, the origin of the O coordinate system is the optical center of the multi-line lidar, the Z-axis is vertical to the road surface and downward is positive, the Y-axis is parallel to the direction of travel of the road roller, and the X-axis is parallel to the wheel axle center line of the road roller.

[0058] Specifically, such as Figure 5 As shown, when filtering point clouds, based on the multi-line lidar installed at a ground clearance height H, two preset height thresholds th1 and th2 are used to extract point clouds with heights between H-th1 and H+th2, filtering out point clouds with heights less than H-th1 and greater than H+th2 to form the point clouds of the road surface and curb stones. For example, when the ground clearance height H = 1250mm, H-th1 and H+th2 can be set to 1100mm and 1350mm respectively.

[0059] Since point clouds have a scanning order, after some point clouds are filtered out, the outlines of the road surface and curb stones will be partially missing. Therefore, the point clouds before and after the filtered point clouds are arranged in order, skipping the missing parts, so that the outlines of the road surface and curb stones are still sharp and distinct.

[0060] This application embodiment considers the presence of large protruding stones, manhole covers, and fluctuations in the radar itself. The point cloud of the area can be processed using the Kalman filter method according to the road surface conditions (this step is not necessary if the road surface preloading is good). The Kalman filter can effectively suppress vertical fluctuations and protrusions while ensuring minimal changes in the horizontal direction, and can also retain the corner information of the road edge.

[0061] S102. Divide the point cloud into multiple first grid cells in the horizontal direction, obtain the center point cloud and point cloud direction of the first grid cell, the horizontal direction is the direction from the road surface to the curb, and the coordinates of the center point cloud of the first grid cell are the average coordinates of the point cloud in the first grid cell.

[0062] In one embodiment, obtaining the center point cloud and point cloud orientation of the first grid cell includes: obtaining the point cloud orientation of the first grid cell using principal component analysis.

[0063] In this embodiment, the horizontal direction refers to the direction from the road surface to the curb, which is also the X-axis direction of the multi-line lidar; the window size of the first grid unit can be set according to the actual situation, such as dividing the first grid unit into 10cm.

[0064] like Figure 6 As shown, the point cloud is divided into horizontal grids to obtain multiple first grid cells. Each first grid cell contains multiple point clouds. The PCA (Principal Component Analysis) method is used to obtain the center point cloud and point cloud orientation of the first grid cell.

[0065] The process of obtaining the center point cloud and point cloud orientation of the first grid cell using PCA principal component analysis is as follows:

[0066] A1: Solve for the average value of the point cloud in the X and Y directions in the O-XYZ coordinate system within the first grid cell, i.e., the center point cloud.

[0067] A2: Construct the covariance matrix for the point cloud within the first grid cell in the O-XYZ coordinate system.

[0068] A3: Solve for the orthogonal normalized vector (ε) of the eigenvectors of the covariance matrix C. x , ε y );

[0069] A4: Using vector (ε) x , ε y ) are the two direction vectors (i.e., the two point cloud directions) of the point cloud within the first grid cell.

[0070] S103. Cluster multiple first grid cells into a curb area and a road surface area according to the center point cloud and the point cloud direction of the first grid cell.

[0071] In one implementation, clustering multiple first grid cells into a curb area and a road surface area according to the center point cloud and the point cloud direction of the first grid cell includes:

[0072] When the height difference between the center point coordinate of the first grid cell and the installation height of the multi-line lidar is less than a preset threshold, the height difference between the center point cloud of the first grid cell and the center point cloud of the previous first grid cell is less than a preset height difference, and the angle difference between the point cloud direction of the first grid cell and the point cloud direction of the previous first grid cell is less than a preset angle difference, cluster the first grid cell into the road surface area;

[0073] When the height difference between the center point cloud of the first grid cell and the center point cloud of the previous first grid cell is greater than a preset height difference, and the angle difference between the point cloud direction of the first grid cell and the point cloud direction of the previous first grid cell is greater than a preset angle difference, cluster the first grid cell into the curb area.

[0074] According to the center point cloud and the point cloud direction of the first grid cell, it can be determined whether the first grid cell is a road surface area or a curb area, and the road surface area and the curb area can be roughly distinguished. As Figure 7 shown, if two adjacent first grid cells have similar characteristics, and the center point coordinate of the first grid cell and the radar installation height H satisfy |y - H| < th, where th is a settable threshold, then it is clustered into the road surface area. To avoid clustering into an area above the curb with similar road surface characteristics, the center coordinate of the first grid cell satisfies |y - H| < th; and two adjacent first grid cells have similar characteristics, such as the height of the center point cloud and the angle change of the point cloud direction of two adjacent first grid cells are not large, that is, the height difference between the center point cloud of the first grid cell and the center point cloud of the previous first grid cell is less than a preset height difference, and the angle difference between the point cloud direction of the first grid cell and the point cloud direction of the previous first grid cell is less than a preset angle difference, then it is clustered into the road surface area.

[0075] If the characteristics of two adjacent first grid cells are quite different, that is, the height difference between the center point cloud of the first grid cell and the center point cloud of the previous first grid cell is greater than a preset height difference, and the angle difference between the point cloud direction of the first grid cell and the point cloud direction of the previous first grid cell is greater than a preset angle difference, then it is clustered into the curb area.

[0076] S104. Obtain a curb point cloud set and a road surface point cloud set according to the curb area and the road surface area.

[0077] In one implementation, obtaining the curb point cloud and the road point cloud based on the curb area and the road surface area includes:

[0078] The curb area is divided into multiple second grid cells in the horizontal direction, and the center point cloud and point cloud direction of the second grid cell are obtained. The size of the second grid cell is smaller than the size of the first grid cell.

[0079] Based on the center point cloud and point cloud direction of the second grid cell, multiple second grid cells are clustered to obtain the curb point cloud set and the road surface point cloud set.

[0080] After clustering the first grid cell to roughly divide the road surface area and the curb area, the curb area not only contains the point cloud of the curb area but may also contain the point cloud of the road surface area. This latter part of the point cloud cannot be directly used as the curb area point cloud set; a fine search is needed to exclude the point cloud of the road surface area. The coarsely located curb area is then divided into horizontal grids, resulting in multiple second grid cells, each smaller than the first grid cell. Principal component analysis (PCA) is used to obtain the center point cloud and point cloud orientation of the second grid cells. Using the center point cloud and point cloud orientation of the second grid cells, the road surface and curb in the curb area can be further refined, resulting in accurate and complete curb and road surface point cloud sets. The PCA method can be referenced from the analysis method for the first grid cell and will not be elaborated here.

[0081] In the point cloud, points within a preset distance threshold from the fitted line of the roadside region are searched, and the roadside point cloud is expanded to obtain the complete point cloud of the roadside region. The purpose of this step is that, for sloping roads, if the boundary point between the road and the roadside is placed at the center of the first grid cell during mesh generation, the main direction may be submerged by the road surface and thus clustered on the ground. This embodiment of the application can perform a reverse search to the complete roadside region. At the same time, the more roadside points there are, the more beneficial it is to the accuracy of the final roadside plane fitting.

[0082] S105. Perform linear fitting on the curb point cloud and the road surface point cloud to obtain the curb fitting line and the road surface fitting line. Based on the intersection of the curb fitting line and the road surface fitting line, select the point cloud that is closest to the intersection point and is in the curb point cloud as the boundary point between the road surface and the curb stone.

[0083] After obtaining the complete point cloud of the curb area, the points with x coordinates from 0 to the curb point cloud are the road surface point cloud. Straight line fitting is performed on the curb point cloud and the road surface point cloud respectively to obtain the curb fitting line and the road surface fitting line. The curb fitting line and the road surface fitting line intersect. The point closest to the intersection point and in the curb point cloud is selected as the boundary point between the road surface and the curb.

[0084] like Figure 8As shown, the point cloud that is both located on the edge point cloud and closest to the intersection of the edge fitting line and the road surface fitting line is selected as the dividing point because if there are protruding stones close to the edge, the traditional method of searching for corner points will likely select the point cloud in front of the stones, resulting in a large calculated edge distance and a large under-pressure area.

[0085] Additionally, if the roadside boundary is not identified in the point cloud within the coordinate range x∈[-x1,x2], no localization result will be output; otherwise, one-sided or two-sided localization results will be output as needed, following the steps described above.

[0086] S106. Based on the dividing point, calculate the edge-to-edge distance and tilt angle of the road roller.

[0087] like Figure 9 As shown, AB is the intersection of the vertical surface of the curbstone and the ground, BC is the intersection of the left side of the roller wheel and the ground, and CD is the intersection of the plane containing the 0° scanning line of the multi-line lidar and the ground. CD is the edge-fitting distance referred to in this embodiment, which is the distance from the left side of the roller wheel to the curbstone; ∠ABC is the pose angle referred to in this embodiment, which is the angle between the direction of travel of the roller wheel and the curbstone.

[0088] After determining the boundary point between the road surface and the curb, the edge-fitting distance can be obtained by subtracting the side distance W between the multi-line lidar and the roller wheel from the x-coordinate of the boundary point. For example, if the coordinates of the point cloud of the boundary point are (388, 1254, 0), and the side distance W between the multi-line lidar and the roller wheel is 380mm, the edge-fitting distance is 388-380=8mm.

[0089] In one implementation, the distance to the edge and the tilt angle of the roller are calculated based on the boundary point, including:

[0090] Plane fitting is performed on the point cloud of the curbstone to obtain the normal vector of the plane where the curbstone is located; based on the normal vector of the plane where the curbstone is located, the tilt angle of the roller is calculated.

[0091] like Figure 10 As shown, the top of the road roller is projected onto the ground to form an OXY plane. The intersection of the curbstone and the ground is the thick solid line in the figure. Setting Z=0, the point cloud of the curbstone corresponding to the thick solid line can be obtained. Performing plane fitting on the point cloud of the curbstone, the thick solid line is obtained as ax+by+d=0, and then the slope of the solid line is calculated. The tilt angle of the pose is

[0092] Specifically, plane fitting can be performed on the point cloud data of the curbstone using the least squares method or the planar random sampling consensus algorithm to obtain the normal vector (a, b, c) of the plane containing the curbstone, and then the tilt angle of the road roller can be calculated. For example, if the normal vector of the plane containing the curbstone is obtained through plane fitting as (0.5126, -0.00096, 0.4883), then the pose tilt angle is...

[0093] A second aspect of this application provides a lidar-based edge positioning device for a road roller, comprising:

[0094] The scanning module is used to acquire point clouds obtained by multi-line lidar scanning of the road surface and curb stones;

[0095] The partitioning module is used to divide the point cloud into multiple first grid cells in the horizontal direction, obtain the center point cloud and point cloud direction of the first grid cell, the horizontal direction is the direction from the road surface to the curb, and the coordinates of the center point cloud of the first grid cell are the average coordinates of the point cloud in the first grid cell.

[0096] The clustering module is used to cluster multiple first grid cells based on the center point cloud and point cloud direction of the first grid cell to obtain the curb area and the road surface area;

[0097] The acquisition module is used to acquire the curb point cloud and the road point cloud based on the curb area and the road surface area;

[0098] The boundary module is used to perform straight line fitting on the curb area and the road surface area to obtain the curb fitting line and the road surface fitting line. Based on the intersection of the curb fitting line and the road surface fitting line, the point cloud closest to the intersection point and in the curb point cloud is selected as the boundary point between the road surface and the curb stone.

[0099] The calculation module is used to calculate the roller's edge-fitting distance and tilt angle based on the boundary point.

[0100] The lidar-based roller edge positioning device of this application corresponds to the method of the first aspect of this application. For specific implementation details, please refer to the method of the first aspect, which will not be repeated here.

[0101] A third aspect of this application provides a roller edge positioning system, including a multi-line lidar and a lidar-based roller edge positioning device as described in the second aspect, wherein the ground clearance H of the multi-line lidar and the side distance W between the multi-line lidar and the side of the roller satisfy the following constraints:

[0102]

[0103] like Figure 11As shown, the height H above the ground refers to the vertical distance from the optical center O of the multi-line lidar to the ground, the lateral distance W refers to the horizontal distance from the optical center O of the multi-line lidar to the leftmost or rightmost side of the road roller, and k is the scanning angle resolution of the multi-line lidar. For example, the scanning angle resolution of the four-line lidar mentioned above is 0.2°.

[0104] like Figure 12 As shown, OO1 is the ground clearance H of the multi-line lidar. Assuming the road roller is very close to the curb, with the dividing point at P1, then O1P1 is the lateral distance W between the multi-line lidar and the road roller. The next adjacent ray, OP2, is projected onto the curb; if there were no curb obstruction, it would be projected onto the road surface at point P'2. To ensure the positioning accuracy of the road roller is controlled within 1cm, the distance between P1 and P'2 must be less than or equal to 1cm, which can be expressed mathematically as: By limiting the constraint relationship between the two, it can be ensured that the lateral distance between two adjacent points on the curb from the road surface scan is within 1cm, which means that the edge-fitting distance can be controlled within 1cm.

[0105] For ease of adjustment, the edge positioning system also includes an adjustable bracket. The installation position of the multi-line lidar can be adjusted by the adjustable bracket. The adjustable bracket mounts the multi-line lidar on the road roller and is used to adjust the ground clearance H of the multi-line lidar and the side distance W between the multi-line lidar and the side of the road roller.

[0106] In this embodiment of the application, the side distance W is greater than or equal to 10cm and less than or equal to half the width of the road roller.

[0107] Considering the possibility of vegetation encroachment on the outer side of the curbstone, in order to prevent the scanning light of the multi-line lidar from being blocked by vegetation during the edge-fitting operation, the side distance W of the multi-line lidar is required to be greater than or equal to 10cm; and the side distance W is required to be less than or equal to half the width of the road roller. That is to say, when the road roller wheel is performing edge-fitting operation on the left side of the curbstone, the side distance W between the multi-line lidar and the left side of the road roller is less than or equal to half the width of the road roller.

[0108] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method as described in any of the first aspects.

[0109] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0110] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0111] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0112] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0113] In the embodiments provided in this application, it should be understood that the disclosed devices / terminal equipment and methods can be implemented in other ways. For example, the device / terminal equipment embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling or direct coupling or communication connection may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0114] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0115] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0116] If the integrated module / unit is implemented as 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, all or part of the processes in the methods of the above embodiments can also be implemented by hardware related to computer program instructions. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electrical carrier signals and telecommunication signals.

[0117] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for positioning a road roller along its edge based on lidar, characterized in that, include: Acquire point clouds obtained from multi-line lidar scanning of the road surface and curb stones; The point cloud is divided into multiple first grid units in the horizontal direction, and the center point cloud and point cloud direction of the first grid unit are obtained. The horizontal direction is the direction from the road surface to the curb stone. The coordinates of the center point cloud of the first grid unit are the average coordinates of the point cloud in the first grid unit. Based on the center point cloud and point cloud direction of the first grid cell, multiple first grid cells are clustered to obtain the roadside area and the road surface area; Based on the curb area and the road surface area, obtain the curb point cloud and the road surface point cloud; Linear fitting is performed on the curb point cloud and the road surface point cloud to obtain the curb fitting line and the road surface fitting line. Based on the intersection of the curb fitting line and the road surface fitting line, the point cloud closest to the intersection point and in the curb point cloud is selected as the boundary point between the road surface and the curb stone. Based on the boundary point, the edge-fitting distance and tilt angle of the road roller are calculated; The step of obtaining the curb point cloud and the road point cloud based on the curb area and the road surface area includes: The roadside area is divided into multiple second grid units in the horizontal direction, and the center point cloud and point cloud direction of the second grid unit are obtained, wherein the second grid unit is smaller than the first grid unit; Based on the center point cloud and point cloud direction of the second grid cell, multiple second grid cells are clustered to obtain the curb point cloud set and the road surface point cloud set.

2. The method for positioning a road roller along its edge based on lidar according to claim 1, characterized in that, The step of clustering multiple first grid cells to obtain the curb region and road surface region based on the center point cloud and point cloud direction of the first grid cell includes: When the height difference between the center point coordinates of the first grid cell and the installation height of the multi-line lidar is less than a preset threshold, the height difference between the center point cloud of the first grid cell and the center point cloud of the previous first grid cell is less than a preset height difference, and the angle difference between the point cloud direction of the first grid cell and the point cloud direction of the previous first grid cell is less than a preset angle difference, the first grid cell is clustered into a road surface area. When the height difference between the center point cloud of the first grid cell and the center point cloud of the previous first grid cell is greater than a preset height difference, and the angle difference between the point cloud direction of the first grid cell and the point cloud direction of the previous first grid cell is greater than a preset angle difference, the first grid cell is clustered into a curb region.

3. The method for positioning a road roller along its edge based on lidar according to claim 1, characterized in that, The process of obtaining the center point cloud and point cloud orientation of the first grid cell includes: The point cloud orientation of the first grid cell is obtained using principal component analysis.

4. The method for positioning a road roller along its edge based on lidar according to claim 1, characterized in that, After acquiring the point cloud obtained from the multi-line lidar scan of the road surface and curbstone, the method further includes: The point cloud is filtered, and the filtered point cloud p is: in, n This refers to the line number of the scanning line of the multi-line lidar. i The point number of the scan line. H The height of the multi-line lidar above the ground. th1 and th2 Preset height threshold ，x , y, z The initial point cloud is in O-XYZ Coordinates of a point in a coordinate system O The origin of the coordinate system is the optical center of the multi-line lidar. Z With the axis vertical and the road surface pointing downwards, that is considered positive. Y The shaft is parallel to the direction of travel of the road roller. X The shaft is parallel to the centerline of the roller wheel axle.

5. The method for positioning a road roller along its edge based on lidar according to claim 1, characterized in that, The step of calculating the edge-fitting distance and tilt angle of the road roller based on the boundary point includes: Plane fitting is performed on the point cloud of the curbstone to obtain the normal vector of the plane where the curbstone is located; The tilt angle of the road roller is calculated based on the normal vector of the plane where the curbstone is located.

6. A roller edge positioning device based on lidar, characterized in that, The method applied to any one of claims 1 to 5 includes: The scanning module is used to acquire point clouds obtained by multi-line lidar scanning of the road surface and curb stones; The partitioning module is used to divide the point cloud into multiple first grid cells in the horizontal direction, obtain the center point cloud and point cloud direction of the first grid cell, wherein the horizontal direction is the direction from the road surface to the curb stone, and the coordinates of the center point cloud of the first grid cell are the average coordinates of the point cloud in the first grid cell. The clustering module is used to cluster multiple first grid cells based on the center point cloud and point cloud direction of the first grid cell to obtain the curb area and the road surface area; The acquisition module is used to acquire the curb point cloud and the road point cloud based on the curb area and the road surface area; The boundary module is used to perform straight line fitting on the curb area and the road surface area to obtain the curb fitting line and the road surface fitting line. Based on the intersection of the curb fitting line and the road surface fitting line, the point cloud closest to the intersection point and in the curb point cloud set is selected as the boundary point between the road surface and the curb stone. The calculation module is used to calculate the edge-fitting distance and tilt angle of the road roller based on the boundary point.

7. A roller edge positioning system, characterized in that, The device includes a multi-line lidar and a lidar-based road roller edge positioning device as described in claim 6, wherein the ground clearance H of the multi-line lidar and the side distance W between the multi-line lidar and the side of the road roller satisfy the following constraint relationship: Where k is the scanning angle resolution of the multi-line lidar.

8. A roller edge positioning system according to claim 7, characterized in that, The system also includes an adjustable bracket connected to the multi-line lidar for adjusting the ground clearance H and the lateral distance W of the multi-line lidar.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 5.

Images