Calibration method of field end laser radar for vehicle-road cooperation

Abstract

The invention belongs to the technical field of automatic driving, and discloses a calibration method of a field end laser radar for vehicle-road cooperation, and the method comprises the steps: obtaining point cloud data through laser radar driving, selecting a point cloud belonging to the ground in a visual interface, and achieving the extraction of a ground part in the point cloud through a plane fitting algorithm. Reversely deriving the attitude and the terrain clearance of the laser radar, determining the longitude, latitude and course angle of the laser radar, re-selecting the point cloud of the lane line from the obtained point cloud, projecting the selected point cloud to the map, rotating and translating the original point of the laser radar until the point cloud is matched with the lane line on the map, and obtaining the specific parameters of translation and rotation. And longitude and latitude and orientation of the laser radar are obtained through longitude and latitude conversion. By adopting the calibration method, the laser radar deployed at the field end can be calibrated, and testers do not need to go to the field to collect data, so that the calibration cost and the calibration time are greatly reduced.

Description

Technical Field

[0001] The present application relates to the field of autonomous driving technology, and more specifically, to a calibration method for a field-side laser radar used for vehicle-road collaboration. Background Art

[0002] Roadside LiDAR is one of the important sensors for obtaining high-quality perception results. The LiDAR itself can sense the exact position of the target in its own coordinate system. In vehicle-road collaboration, the LiDAR coordinate system needs to be transferred to the world coordinate system before it can be used by the vehicle. Therefore, it is necessary to accurately calibrate the position (latitude, longitude and posture) of each LiDAR so that the roadside perception results can be effectively utilized.

[0003] The existing lidar calibration methods have the following main defects, including:

[0004] 1. Only applicable to indoors: The current LiDAR calibration algorithm requires a closed environment to be set up indoors. Special targets (such as QR codes, objects of regular shapes, etc.) are placed in the closed environment. The LiDAR is calibrated by identifying the targets. However, the field LiDAR is deployed outdoors or in public parking lots. The environment changes greatly and it is impossible to deploy a unified calibration environment.

[0005] 2. Testers are required to go to the site to mark points: Currently, field-side LiDAR calibration requires testers to carry RTK handheld base stations to the site to collect precise positioning points, or carry special targets and place them on site for LiDAR to identify. This solution not only has huge labor costs, but also a long calibration cycle. Summary of the invention

[0006] In order to solve the above problems, the present application provides a calibration method for a field-side laser radar for vehicle-road collaboration.

[0007] The present application provides a calibration method for a field-side laser radar for vehicle-road collaboration using the following technical solution:

[0008] A calibration method for a field-side laser radar for vehicle-road collaboration includes the following steps:

[0009] S1. Install the field-side laser radar and start the calibration procedure;

[0010] S2, driving through the laser radar to obtain point cloud data;

[0011] S3. In the visualization interface, select the point cloud belonging to the ground;

[0012] S4, extracting the ground part in the point cloud through plane fitting algorithm;

[0013] S5. Reversely derive the attitude and height above the ground of the laser radar;

[0014] S6, determine the latitude and longitude and heading angle of the laser radar, and select the point cloud of the lane line again from the point cloud obtained in S2;

[0015] S7, projecting the point cloud selected in S6 onto the map, and rotating and translating the laser radar origin until the point cloud matches the lane line on the map;

[0016] S8. Obtain the specific parameters of translation and rotation in S7, and obtain the latitude, longitude and orientation of the laser radar through longitude and latitude conversion, and the calibration is completed.

[0017] Furthermore, in S3, no less than 10 points of the point cloud belonging to the ground are selected.

[0018] Furthermore, the plane fitting algorithm in S4 includes the following steps:

[0019] S401, sorting the selected point clouds by z-axis size;

[0020] S402, selecting the average Z value of several points with the smallest z-axis;

[0021] S403, setting a height threshold, and taking a point set consisting of points near the threshold as a seed point set;

[0022] S404, using a linear model for plane model estimation, where the formula for fitting the ground equation is: ax+by+cz+d=0, that is, n T x = -d;

[0023] S405, calculate the covariance matrix C and the singular vector of the minimum singular value to solve n, that is, n = [a, b, c] T ;

[0024] S406, determine a plane through n, and use the seed point set as the initial point set, and its covariance matrix is:

[0025] S407, calculating the orthogonal projection distance from each point in the point cloud to the plane, and comparing the distance with a set threshold to determine the ground point;

[0026] S408: Treat all classified ground points as the next seed point set for iterative optimization to obtain a result.

[0027] In summary, the present application includes at least one of the following beneficial technical effects:

[0028] The present invention can accurately calibrate the laser radar sensor installed outdoors, including external parameters such as the position and direction of the laser radar in the world coordinate system, so that the point cloud data collected by the laser radar can be accurately represented in the world coordinate system for use in subsequent vehicle-road collaborative tasks; this calibration method can be used to calibrate the laser radar deployed on the field; and this calibration method does not require testers to collect data on site, which greatly reduces the calibration cost and calibration time. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is a flowchart of the present application. DETAILED DESCRIPTION

[0030] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application; obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments, and all other embodiments obtained by ordinary technicians in this field based on the embodiments in the present application without making creative work are within the scope of protection of the present application.

[0031] In the description of the present application, it should be noted that the terms "upper", "lower", "inner", "outer", "top / bottom" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific position, be constructed and operated in a specific position, and therefore cannot be understood as limiting the present application. In addition, the terms "first" and "second" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0032] In the description of this application, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "provided with", "set / connected", "connected", etc. should be understood in a broad sense. For example, "connected" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0033] Embodiment 1:

[0034] The present application embodiment discloses a calibration method for a field-side laser radar for vehicle-road collaboration, comprising the following steps:

[0035] S1. Install the field-side laser radar and start the calibration procedure;

[0036] S2, driving through the laser radar to obtain point cloud data;

[0037] S3. In the visualization interface, select the point cloud belonging to the ground;

[0038] S4, extracting the ground part in the point cloud through plane fitting algorithm;

[0039] S5. Reversely derive the attitude and height above the ground of the laser radar;

[0040] S6, determine the latitude and longitude and heading angle of the laser radar, and select the point cloud of the lane line again from the point cloud obtained in S2;

[0041] S7, projecting the point cloud selected in S6 onto the map, and rotating and translating the laser radar origin until the point cloud matches the lane line on the map;

[0042] S8. Obtain the specific parameters of translation and rotation in S7, and obtain the latitude, longitude and orientation of the laser radar through longitude and latitude conversion, and the calibration is completed.

[0043] Preferably, at least 10 points of the point cloud belonging to the ground are selected in S3.

[0044] Preferably, the plane fitting algorithm in S4 comprises the following steps:

[0045] S401, sorting the selected point clouds by z-axis size;

[0046] S402, selecting the average Z value of several points with the smallest z-axis;

[0047] S403, setting a height threshold, and taking a point set consisting of points near the threshold as a seed point set;

[0048] S404, using a linear model for plane model estimation, where the formula for fitting the ground equation is: ax+by+cz+d=0, that is, n T x = -d;

[0049] S405, calculate the covariance matrix C and the singular vector of the minimum singular value to solve n, that is, n = [a, b, c] T ;

[0050] S406, determine a plane through n, and use the seed point set as the initial point set, and its covariance matrix is:

[0051] S407, calculating the orthogonal projection distance from each point in the point cloud to the plane, and comparing the distance with a set threshold to determine the ground point;

[0052] S408: Treat all classified ground points as the next seed point set for iterative optimization to obtain a result.

[0053] The present invention can accurately calibrate the laser radar sensor installed outdoors, including external parameters such as the position and direction of the laser radar in the world coordinate system, so that the point cloud data collected by the laser radar can be accurately represented in the world coordinate system for use in subsequent vehicle-road collaborative tasks; this calibration method can be used to calibrate the laser radar deployed on the field; and this calibration method does not require testers to collect data on site, which greatly reduces the calibration cost and calibration time.

[0054] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A calibration method for a field-side laser radar for vehicle-road collaboration, characterized in that: The following steps are involved: S1. After the laser radar is installed on the field, the calibration procedure begins; S2, driving through the laser radar to obtain point cloud data; S3. In the visualization interface, select the point cloud belonging to the ground; S4, extracting the ground part in the point cloud through plane fitting algorithm; S5. Reversely derive the attitude and height above the ground of the laser radar; S6, determine the latitude and longitude and heading angle of the laser radar, and select the point cloud of the lane line again from the point cloud obtained in S2; S7, projecting the point cloud selected in S6 onto the map, and rotating and translating the laser radar origin until the point cloud matches the lane line on the map; S8. Obtain the specific parameters of translation and rotation in S7, and obtain the latitude, longitude and orientation of the laser radar through longitude and latitude conversion, and the calibration is completed.

2. The calibration method of a field-side laser radar for vehicle-road collaboration according to claim 1 is characterized in that: In S3, no less than 10 points of the point cloud belonging to the ground are selected.

3. The calibration method of a field-side laser radar for vehicle-road collaboration according to claim 1 is characterized in that: The plane fitting algorithm in S4 includes the following steps: S401, sorting the selected point clouds by z-axis size; S402, selecting the average Z value of several points with the smallest z-axis; S403, setting a height threshold, and taking a point set consisting of points near the threshold as a seed point set; S404, using a linear model for plane model estimation, where the formula for fitting the ground equation is: ax+by+cz+d=0, that is, n T x = -d; S405, calculate its covariance matrix C and the singular vector of the minimum singular value to solve n, that is, n = [a, b, c] T ; S406, determine a plane through n, and use the seed point set as the initial point set, and its covariance matrix is: S407, calculating the orthogonal projection distance from each point in the point cloud to the plane, and comparing the distance with a set threshold to determine the ground point; S408: Treat all classified ground points as the next seed point set for iterative optimization to obtain a result.

Images