Radar calibration method and device, electronic equipment and storage medium
By acquiring and processing the point clouds collected by the target radar, determining the intersection points of the target and performing radar calibration, the problems of environmental dependence and low accuracy in the existing technology are solved, and high-precision radar calibration is achieved.
Patent Information
- Application Number
- CN202311595896.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-11-27
- Publication Date
- 2025-06-03
AI Technical Summary
The existing radar calibration technology has limitations and needs to be carried out in a specific environment, increasing the cost of calibration and affecting accuracy.
By acquiring the target point cloud about the preset area collected by the target radar of the target vehicle, the point cloud corresponding to the first panel and the second panel is determined, the intersection point between the target and the ground is calculated, and the target radar is calibrated based on this point.
The feature extraction accuracy is improved, the accuracy of radar calibration is ensured, and the problem of inability to accurately extract feature points is solved.
Smart Images
Figure CN120085264A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the technical field of radar calibration. Specifically, it relates to a method, device, electronic device, and storage medium for radar calibration. Background Art
[0002] Radar calibration refers to the precise calibration of a radar to ensure its ability to accurately measure the distance, angle, and position of a target object. Radar calibration is a very important part of intelligent driving technology, which directly affects the measurement accuracy and reliability of the radar.
[0003] In the prior art, when performing radar calibration, it often relies on a target. Common targets mainly include passive reflection targets, dynamic targets, uniform reflection targets, and distance plates, etc. However, when using these targets in the prior art, there are certain limitations, such as requiring a specific environment, etc. These factors not only increase the cost of calibration but also have a certain impact on the calibration accuracy. Summary of the Invention
[0004] In view of this, the purpose of this application is to provide a method, device, electronic device, and storage medium for radar calibration to overcome the problems in the prior art.
[0005] In a first aspect, an embodiment of this application provides a method for radar calibration, and the method includes:
[0006] Obtain target point clouds of a preset area collected by a target radar of a target vehicle, where a preset number of targets are set in the preset area, and the target includes a first panel and a second panel at a preset angle;
[0007] Based on the target point clouds, determine first panel point clouds and second panel point clouds corresponding to the first panel and the second panel respectively;
[0008] Based on the panel intersection line of the first panel point clouds and the second panel point clouds, determine the intersection point position of the target and the ground;
[0009] Calibrate the target radar based on the intersection point position and a reference intersection point position corresponding to the intersection point position.
[0010] In some technical solutions of this application, the above-mentioned determining, based on the target point clouds, first panel point clouds and second panel point clouds corresponding to the first panel and the second panel respectively includes:
[0011] Perform ground correction on the target point clouds, filter out some point clouds in the target point clouds, and obtain remaining point clouds;
[0012] Perform clustering processing on the remaining point cloud to obtain the first panel point cloud corresponding to the first panel and the second panel point cloud corresponding to the second panel.
[0013] In some technical solutions of the present application, the panel intersection line of the above-mentioned first panel point cloud and the second panel point cloud is determined by the following method:
[0014] Determine the intersection line vector according to the first normal vector of the first panel and the second normal vector of the second panel;
[0015] Calculate the intersection point position of the first panel and the second panel according to the first plane equation of the first panel and the second plane equation of the second panel;
[0016] Determine the panel intersection line of the first panel point cloud and the second panel point cloud according to the intersection line vector and the intersection point position of the first panel and the second panel.
[0017] In some technical solutions of the present application, the above-mentioned calculating the intersection point position of the first panel and the second panel according to the first plane equation of the first panel and the second plane equation of the second panel includes:
[0018] Construct an augmented matrix based on the first plane equation and the second plane equation;
[0019] Perform Gaussian elimination on the augmented matrix to determine the intersection point position of the first panel and the second panel.
[0020] In some technical solutions of the present application, the above-mentioned determining the panel intersection line of the first panel point cloud and the second panel point cloud according to the intersection line vector and the intersection point position of the first panel and the second panel includes:
[0021] Determine the transformation ratio based on the coordinate component of the intersection line vector along the preset direction and the coordinate value of the intersection point position in the preset direction;
[0022] Determine the panel intersection line of the first panel point cloud and the second panel point cloud based on the transformation ratio and the line segment model.
[0023] Second, an embodiment of the present application provides a radar calibration device, and the device includes:
[0024] An acquisition module, configured to acquire target point cloud about a preset area collected by a target radar of a target vehicle, where a preset number of targets are set in the preset area, and the targets include a first panel and a second panel at a preset angle;
[0025] The first determination module is configured to determine a first panel point cloud and a second panel point cloud corresponding to the first panel and the second panel respectively based on the target point cloud;
[0026] The second determination module is configured to determine the intersection position of the target and the ground based on the panel intersection line of the first panel point cloud and the second panel point cloud;
[0027] The calibration module is configured to calibrate the target radar based on the intersection position and a reference intersection position corresponding to the intersection position.
[0028] In some technical solutions of the present application, when determining the first panel point cloud and the second panel point cloud corresponding to the first panel and the second panel respectively based on the target point cloud, the above-mentioned first determination module is configured to:
[0029] Perform ground correction on the target point cloud, filter out some point clouds in the target point cloud, and obtain a remaining point cloud;
[0030] Perform clustering processing on the remaining point cloud to obtain the first panel point cloud corresponding to the first panel and the second panel point cloud corresponding to the second panel.
[0031] In some technical solutions of the present application, the above-mentioned second determination module determines the panel intersection line of the first panel point cloud and the second panel point cloud in the following manner:
[0032] Determine an intersection line vector according to the first normal vector of the first panel and the second normal vector of the second panel;
[0033] Calculate the intersection position of the first panel and the second panel according to the first plane equation of the first panel and the second plane equation of the second panel;
[0034] Determine the panel intersection line of the first panel point cloud and the second panel point cloud according to the intersection line vector and the intersection position of the first panel and the second panel.
[0035] In a third aspect, an embodiment of the present application provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the steps of the above-mentioned radar calibration method are implemented.
[0036] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is run by a processor, the steps of the above-mentioned radar calibration method are executed.
[0037] The technical solutions provided by the embodiments of the present application may include the following beneficial effects:
[0038] The method of the present application includes obtaining target point clouds of a preset area collected by a target radar of a target vehicle, wherein a preset number of targets are set in the preset area, and the target includes a first panel and a second panel at a preset angle; based on the target point clouds, determining a first panel point cloud and a second panel point cloud corresponding to the first panel and the second panel respectively; based on the panel intersection line of the first panel point cloud and the second panel point cloud, determining the intersection point of the target and the ground; and calibrating the target radar based on the intersection point and a reference intersection point corresponding to the intersection point. The present application calibrates the radar through a target including a first panel and a second panel, which not only solves the problem of inability to accurately extract feature points, but also improves the feature extraction accuracy and ensures the accuracy of radar calibration.
[0039] To make the above objects, features, and advantages of the present application more obvious and understandable, the following specifically enumerates preferred embodiments and, in conjunction with the accompanying drawings, details are described as follows. BRIEF DESCRIPTION OF THE DRAWINGS
[0040] To more clearly illustrate the technical solutions of the embodiments of the present application, the following briefly introduces the drawings required for the embodiments. It should be understood that the following drawings only show some embodiments of the present application and should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.
[0041] Figure 1 Shows a schematic flowchart of a method for radar calibration provided by an embodiment of the present application;
[0042] Figure 2 Shows a schematic diagram of a target provided by an embodiment of the present application;
[0043] Figure 3 Shows a schematic diagram of a specific implementation manner provided by an embodiment of the present application;
[0044] Figure 4 Shows a schematic diagram of a device for radar calibration provided by an embodiment of the present application;
[0045] Figure 5 Shows a schematic structural diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0046] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the following will clearly and completely describe the technical solutions in the embodiments of this application with reference to the accompanying drawings in the embodiments of this application. It should be understood that the accompanying drawings in this application are only for the purposes of illustration and description, and are not used to limit the protection scope of this application. Additionally, it should be understood that the schematic drawings are not drawn to scale. The flowcharts used in this application illustrate the operations implemented according to some embodiments of this application. It should be understood that the operations in the flowchart may not be implemented in sequence, and steps without a logical context relationship may be reversed or implemented simultaneously. Moreover, those skilled in the art can add one or more other operations to the flowchart or remove one or more operations from the flowchart under the guidance of the content of this application.
[0047] In addition, the described embodiments are only some embodiments of this application, rather than all of the embodiments. The components of the embodiments of this application usually described and illustrated in the accompanying drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents the selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative efforts belong to the protection scope of this application.
[0048] It should be noted that the term "including" will be used in the embodiments of this application to indicate the existence of the subsequently stated features, but does not exclude the addition of other features.
[0049] Radar calibration refers to the precise calibration of a radar to ensure that it can accurately measure the distance, angle, and position of a target object. Radar calibration is a very important part of intelligent driving technology, which directly affects the measurement accuracy and reliability of the radar.
[0050] When performing lidar calibration in the prior art, it often relies on targets. Common targets mainly include passive reflection targets, dynamic targets, uniform reflection targets, and distance boards, etc. Passive reflection targets: These targets use special materials or reflective coatings to reflect laser beams. They can be planar or three-dimensional shapes, such as spheres, cubes, or prisms. Passive reflection targets are suitable for testing performance indicators such as the ranging, angular resolution, and reflectivity of lidar. Dynamic targets: Dynamic targets are devices that can generate echoes during movement. They simulate moving targets in the real world, such as vehicles, pedestrians, or other obstacles. Dynamic targets can be used to evaluate the performance of lidar in aspects such as target detection, tracking, and measurement accuracy. Uniform reflection targets: These targets have uniform reflection characteristics and can be used to verify the uniformity and sensitivity of lidar. They are usually made of materials with high reflectivity. Distance boards: Distance boards are targets used to calibrate the distance measurement of lidar. They are usually composed of different surface materials and reflection levels and are used to generate echoes at known distances.
[0051] However, when using these targets in the prior art, there are also some drawbacks. The following are some common drawbacks of lidar targets:
[0052] Limited representativeness: Lidar targets have certain limitations when simulating the real world. They may not be able to fully replicate various complex environmental conditions, different types of targets, or movement patterns in real scenarios. Therefore, when using targets for testing, certain performances may not be comprehensively evaluated.
[0053] Dependence on environmental conditions: Some targets may have a certain dependence on specific environmental conditions. For example, passive reflection targets may require appropriate lighting conditions to effectively reflect laser beams. This means that in different environmental conditions, the performance of the targets may vary, thus affecting the evaluation of lidar systems.
[0054] Limited dynamic range: Some dynamic targets may not be able to fully simulate various movement patterns and speeds of real targets. This may lead to inaccurate performance evaluation of lidar in target tracking and motion detection.
[0055] Difficult calibration and maintenance: Lidar targets need to be calibrated and maintained regularly to ensure the accuracy of their reflection characteristics and performance. This requires professional knowledge and equipment, increasing costs and workload.
[0056] Cost and availability: Some lidar targets may be relatively expensive and may not be easily accessible to some research institutions or small enterprises. In addition, some special types of targets may only be available from specific suppliers, with limited availability.
[0057] Poor adaptability of low-beam lidar: The low resolution of low-beam lidar results in a large spacing between laser beams. When the target is placed at a long distance, the target cannot extract complete features in the beam gaps, leading to a decrease in target accuracy.
[0058] Based on this, the embodiments of the present application provide a method, device, electronic device, and storage medium for lidar calibration, which will be described below through embodiments.
[0059] Figure 1 The flowchart of a method for lidar calibration provided by the embodiments of the present application is shown, where the method includes steps S101 - S104; specifically:
[0060] S101. Obtain the target point cloud of a preset area collected by a target radar of a target vehicle, where a preset number of targets are set in the preset area, and the target includes a first panel and a second panel at a preset angle;
[0061] S102. Based on the target point cloud, determine the first panel point cloud and the second panel point cloud corresponding to the first panel and the second panel respectively;
[0062] S103. Based on the panel intersection line of the first panel point cloud and the second panel point cloud, determine the intersection point of the target and the ground;
[0063] S104. Calibrate the target radar based on the intersection point and the reference intersection point corresponding to the intersection point.
[0064] The present application calibrates the radar through a target including a first panel and a second panel, which not only solves the problem of inaccurate extraction of feature points, but also improves the feature extraction accuracy and ensures the accuracy of lidar calibration.
[0065] Some embodiments of the present application will be described in detail below. Without conflict, the following embodiments and the features in the embodiments can be combined with each other.
[0066] Before introducing the lidar calibration method in the embodiments of the present application, the application scenario of the embodiments of the present application will be introduced first. The radar here is the radar on the target vehicle, such as a low-beam radar, etc. Low-beam lidar usually uses a fixed lidar, and only a few laser beams are emitted from the radar scanning head. Once the laser beam touches an object, it will be reflected from the object and form a point cloud on the radar scanning head receiver. Its principle is similar to that of a monocular camera, and 3D information is obtained through multiple 2D images.
[0067] When calibrating the above radar, the embodiments of the present application need to use such as Figure 2The target shown here. The target includes panel 1 and panel 2, and panel 1 and panel 2 are connected by a rotatable device, that is, the angle between panel 1 and panel 2 is adjustable. The angle between panel 1 and panel 2 can be adjusted according to actual measurement needs. In the embodiments of the present application, either of panel 1 and panel 2 can be used as the first panel, and the other can be used as the second panel. In a specific embodiment, the target can be a PVC hard panel with a length of 120 cm and a height of 80 cm. The target can be placed 7-8 meters in front of the radar. It can be understood that in actual operation, different sizes and / or different materials of panels can be selected according to requirements. After setting up the target, the radar on the vehicle starts to be calibrated.
[0068] S101. Obtain the target point cloud of a preset area collected by the target radar of the target vehicle, where a preset number of targets are set in the preset area, and the target includes a first panel and a second panel at a preset angle.
[0069] The vehicle in the embodiments of the present application is an intelligent driving vehicle. The vehicle to be calibrated is called the target vehicle, and the radar set on the target vehicle is called the target radar. The target radar can perform point cloud acquisition. The point cloud collected by the target radar in the embodiments of the present application is called the target point cloud. Since the embodiments of the present application need to use the target to calibrate the radar, the target should be set at a position where the target radar can collect. That is to say, the target point cloud collected by the target radar contains the point cloud of the target.
[0070] S102. Based on the target point cloud, determine the first panel point cloud and the second panel point cloud corresponding to the first panel and the second panel respectively.
[0071] After obtaining the target point cloud collected by the target radar, the embodiments of the present application need to process the target point cloud, and screen out the first panel point cloud of the first panel and the second panel point cloud of the second panel from the target point cloud.
[0072] When processing the target point cloud, it mainly includes: filtering out some point clouds in the target point cloud to obtain the remaining point cloud; performing clustering processing on the remaining point cloud to obtain the first panel point cloud corresponding to the first panel and the second panel point cloud corresponding to the second panel.
[0073] Specifically, the processing of the target point cloud includes: converting the target point cloud into a converted point cloud based on the initial external parameters between the target radar and the ground; screening the converted point cloud according to the preset height requirement to obtain the candidate point cloud; filtering out some point clouds in the target point cloud by performing ground correction on the candidate point cloud to obtain the remaining point cloud.
[0074] Among them, according to the preset height requirement, the converted point cloud is screened to obtain candidate point clouds, including: screening the converted point cloud based on the first height threshold to obtain a preliminary selection of point clouds; screening the preliminary selection of point clouds based on the first height interval to obtain candidate point clouds within the first height interval. The first height threshold is determined based on the initial extrinsic parameters between the target radar and the ground; the first height interval is determined based on the height information of the preliminary selection of point clouds.
[0075] By performing ground correction on the candidate point clouds, some of the point clouds in the target point clouds are filtered out to obtain remaining point clouds, including: calculating the actual ground normal vector of the candidate point clouds; determining the preliminary adjustment extrinsic parameters corresponding to the candidate point clouds based on the difference between the actual ground normal vector and the ideal ground normal vector; iteratively adjusting the candidate point clouds based on the preliminary adjustment extrinsic parameters until the ground height of the adjusted candidate point clouds is less than the preset second height threshold to obtain the adjustment extrinsic parameters, and further obtaining the remaining point clouds.
[0076] Determining the adjustment extrinsic parameters corresponding to the candidate point clouds based on the difference between the actual ground normal vector and the ideal ground normal vector, including: determining the adjustment roll angle and the adjustment pitch angle based on the actual ground normal vector and the ideal ground normal vector; adjusting the candidate point clouds based on the adjustment roll angle and the adjustment pitch angle so that the angle between the adjusted candidate point clouds and the ground is less than the preset angle threshold; calculating the adjusted height of the adjusted candidate point clouds from the ground; taking the adjustment roll angle, the adjustment pitch angle, and the adjusted height as the adjustment extrinsic parameters.
[0077] In the specific implementation of the above analysis process, it can refer to the following Figure 3 way: After obtaining the target point clouds collected by the target radar, the embodiments of the present application need to convert them from the target radar coordinate system to the ground coordinate system. The specific conversion method is: performing coordinate conversion on the target point clouds according to the initial extrinsic parameters of the target radar to obtain the converted point clouds in the ground coordinate system (i.e., the base ground coordinate system). The initial extrinsic parameters here can be determined during the production process of the target vehicle, and they represent the relative position relationship between the radar and the ground after the production of the target vehicle. For example, according to the initial extrinsic parameter T1_guess of the target radar, the target point cloud P1_lidar1 is converted to the converted point cloud P1_base-ground in the ground coordinate system: P1_base-ground = T1_guess * P1_lidar1.
[0078] After obtaining the converted point clouds, in order to improve the partitioning efficiency, the embodiments of the present application first screen the converted point clouds to exclude the height abnormal points in the converted point clouds to obtain candidate point clouds. Here, the basis for excluding the height abnormal points is whether they meet the preset height requirements. If a point cloud meets the height requirements, the point cloud is retained; if a point cloud does not meet the height requirements, the point cloud is excluded.
[0079] The height requirements here include two sub-requirements, namely the height threshold requirement and the height interval requirement. The specific screening process is as follows: Based on the height threshold, the converted point cloud is screened to obtain the preliminary selected point cloud; based on the height interval, the preliminary selected point cloud is screened to obtain the candidate point cloud located within the height interval. In the embodiment of the present application, the point cloud greater than or equal to the height threshold is considered as the abnormal point cloud. Therefore, it is necessary to exclude the point cloud exceeding the height threshold from the converted point cloud to obtain the preliminary selected point cloud with a height less than the height threshold. After obtaining the preliminary selected point cloud, it is necessary to screen the preliminary selected point cloud again. The screening of the preliminary selected point cloud is based on the height interval, and the point cloud not within the height interval in the preliminary selected point cloud is excluded, and the point cloud within the height interval is used as the candidate point cloud.
[0080] In a specific embodiment, the height threshold can be determined based on the initial extrinsic parameters of the target radar. For example, the initial extrinsic parameters of the target radar are multiplied by a preset extrinsic parameter magnification factor to obtain the height threshold. In a specific embodiment, the height interval can be determined based on the height information of each preliminary selected point cloud. For example, the average height of each preliminary selected point cloud is calculated, and the preset height is added to and subtracted from the average height to obtain the height interval. Specifically, the converted point cloud P1_base-ground is sorted by height z, and the points (preliminary selected point cloud) less than 1.5 times (preset extrinsic parameter magnification factor) of the initial extrinsic parameter T1_guess are obtained. The average value of z is calculated, and the range of z average value ± 10 cm (preset height) is used as the candidate point cloud.
[0081] After obtaining the candidate point cloud, it is necessary to identify the ground point cloud and non-ground point cloud in the candidate point cloud. In a specific implementation, the method of performing ground correction on the candidate point cloud is adopted. Specifically, in the embodiment of the present application, the actual ground normal vector of the candidate point cloud is first calculated. For example, the actual ground normal vector of the candidate point cloud is solved by SVD. Then, aiming at being close to the ideal ground normal vector (for example, the normal vector is (0, 0, 1)), the actual ground normal vector of the candidate point cloud is adjusted based on the adjusted extrinsic parameters. Specifically, based on the actual ground normal vector and the ideal ground normal vector, the roll angle and pitch angle are determined; based on the roll angle and the pitch angle, the candidate point cloud is adjusted so that the angle between the adjusted candidate point cloud and the ground is less than the preset angle threshold; the ground height of the adjusted candidate point cloud from the ground is calculated; the roll angle, the pitch angle, and the ground height are used as the adjusted extrinsic parameters.
[0082] When specifically adjusting the actual ground normal vector of the candidate point cloud, continuous iteration is required until the ground height of the adjusted point cloud is less than the preset target height threshold. The target height threshold here is set based on the actual height of the ground. The fact that the ground height of the adjusted point cloud is less than the preset target height threshold indicates that the actual ground normal vector of the candidate point cloud coincides or is close to the ideal ground normal vector (the difference is small and can be ignored), which can also be understood as the candidate point cloud coincides with the actual ground. At this time, the adjusted candidate point cloud is screened again through the second height interval, and the non-ground point cloud is obtained from the candidate point cloud. The embodiments of the present application consider the points in the second height interval in the candidate point cloud as the ground point cloud, and the non-ground point cloud in the candidate point cloud can be obtained after excluding the ground point cloud in the candidate point cloud. The second height interval here is determined based on the adjusted external parameters of the target radar. That is, by adding and subtracting the preset height to and from the adjusted external parameters of the target radar respectively, the second height interval can be obtained.
[0083] In specific implementation, SVD (Singular Value Decomposition, which is used to decompose a matrix into three independent matrices) is used to solve the actual ground normal vector normal1_svd and height h1_svd of the candidate point cloud. Based on the actual ground normal vector normal1_svd and the ideal ground normal vector normal_ground(0, 0, 1), the difference is calculated, and the difference is converted into the roll angle roll1 and pitch angle pitch1. The adjusted variables of the roll angle roll and pitch angle pitch are calculated, and the ground point cloud is adjusted using the variables to make the ground point cloud conform to the real ground. The average height H1_ground - everage of the converted ground point cloud is calculated, and the lidar height ground height H1_fix = -H1_ground - everage is obtained. The adjusted roll1, pitch1 and H1_fix are used as the adjusted external parameter T1_plane - fit for feedback, and the input point cloud is adjusted until the adjusted value of H1_fix is less than 0.5 cm. (The point cloud is pre - processed based on the adjusted external parameter T1_plane - fit and transformed to make the point cloud more parallel to the ground. A new adjusted external parameter T1_plane - fit will be optimized in each iteration). After the ground plane correction, part of the point cloud is filtered out from the target point cloud to obtain the remaining point cloud. In specific implementation, the filtered - out point cloud is generally the point cloud 10 cm above the ground and the point cloud 10 cm below the ground.
[0084] The obtained remaining point cloud contains the calibration target point cloud and other point clouds, so it is also necessary to obtain the calibration target point cloud from the remaining point cloud.
[0085] In order to obtain the target electric cloud, the embodiments of the present application perform clustering processing on the remaining point cloud. In specific implementation, the distance clustering method DBSCAN can be selected. The remaining point cloud is processed by the DBSCAN algorithm to obtain the target point cloud. Since the target in the embodiments of the present application includes a first panel and a second panel, after obtaining the target point cloud, it is also necessary to distinguish the first panel point cloud of the first panel and the second panel point cloud of the second panel.
[0086] S103. Determine the intersection point of the target and the ground based on the panel intersection line of the first panel point cloud and the second panel point cloud.
[0087] After obtaining the first panel point cloud and the second panel point cloud, it is also necessary to determine the panel intersection line of the first panel point cloud and the second panel point cloud. The panel intersection line is determined according to the intersection line vector and the intersection point of the first panel and the second panel. In order to determine the panel intersection line, it is necessary to determine the intersection line vector and the intersection point of the first panel and the second panel.
[0088] The intersection line vector is determined by the first normal vector of the first panel and the second normal vector of the second panel. By extracting the first normal vector of the first panel and the second normal vector of the second panel, the intersection line vector can be obtained. Specifically, the Ransac algorithm is used to fit the target point cloud, and the respective normal vectors of the two surfaces in the target are extracted and denoted as Normal1 and Normal2. Normal1 cross Normal2 = Normal_line, and Normal_line is the direction of the panel intersection line vector. Among them, Normal_line.x, Normal_line.y, and Normal_line.z respectively represent the component magnitudes of the intersection line vector in the x, y, and z dimensions.
[0089] In order to obtain the intersection point of the first panel and the second panel, the embodiments of the present application need to first determine the first surface equation of the first panel and the second surface equation of the second panel. Then, an augmented matrix is constructed using the first surface equation and the second surface equation; after that, Gaussian elimination is performed on the augmented matrix to determine the intersection point of the first panel and the second panel. Specifically, the first surface equation is plane_equation1, and the second surface equation is plane_equation2. The two surface equations are constructed into an augmented matrix, and Gaussian elimination is performed on the augmented matrix to obtain the intersection point P_cross of the two planes, where P_cross.x, P_cross.y, and P_cross.z respectively represent the x, y, and z coordinate values of this point.
[0090] After determining the intersection line vector and the intersection point of the first panel and the second panel, a transformation ratio is determined based on the coordinate component of the intersection line vector along the preset direction and the coordinate value of the intersection point in the preset direction. Then, based on the transformation ratio and the line segment model, the panel intersection line is determined.
[0091] In specific implementation, the transformation ratio t = (0 - P_cross.z) / Normal_line.z. Then, through the transformation ratio t, substituting it into the line segment formula, the x and y positions of the point where the panel intersection line is at a height of 0 are obtained, denoted as P_x and P_y. Combining P_x, P_y, and 0, the position P_target of the intersection point of the target and the ground can be obtained.
[0092] S104. Calibrate the target radar based on the intersection point and the reference intersection point corresponding to the intersection point.
[0093] The reference intersection point corresponding to the intersection point here is set in advance and can be determined based on the center of the target (or other reference points, which can be designed according to actual needs). Specifically, the center position of the target in the ground coordinate system is used as the reference intersection point corresponding to the intersection point of the target and the ground. After determining the intersection point and the reference intersection point corresponding to the intersection point, a registration algorithm can be used to process the two to obtain a registration transformation matrix, and then the target radar can be calibrated based on this registration transformation matrix. For example, the external parameters between the target radar and other vehicle-mounted devices (such as vehicle-mounted inertial navigation devices) are calibrated.
[0094] Figure 4 The structural schematic diagram of a radar calibration device provided by an embodiment of the present application is shown. The device includes:
[0095] An acquisition module, configured to acquire target point clouds of a preset area collected by a target radar of a target vehicle, where a preset number of targets are set in the preset area, and the target includes a first panel and a second panel at a preset angle;
[0096] A first determination module, configured to determine a first panel point cloud and a second panel point cloud corresponding to the first panel and the second panel respectively based on the target point clouds;
[0097] A second determination module, configured to determine the intersection point of the target and the ground based on the panel intersection line of the first panel point cloud and the second panel point cloud;
[0098] A calibration module, configured to calibrate the target radar based on the intersection point and the reference intersection point corresponding to the intersection point.
[0099] Determining the first panel point cloud and the second panel point cloud corresponding to the first panel and the second panel respectively based on the target point cloud includes:
[0100] Perform ground correction on the target point cloud, filter out some point clouds in the target point cloud, and obtain the remaining point cloud;
[0101] Perform clustering processing on the remaining point cloud to obtain the first panel point cloud corresponding to the first panel and the second panel point cloud corresponding to the second panel.
[0102] The panel intersection line of the first panel point cloud and the second panel point cloud is determined by the following method:
[0103] Determine the intersection line vector according to the first normal vector of the first panel and the second normal vector of the second panel;
[0104] Calculate the intersection point position of the first panel and the second panel according to the first plane equation of the first panel and the second plane equation of the second panel;
[0105] Determine the panel intersection line of the first panel point cloud and the second panel point cloud according to the intersection line vector and the intersection point position of the first panel and the second panel.
[0106] Calculating the intersection point position of the first panel and the second panel according to the first plane equation of the first panel and the second plane equation of the second panel includes:
[0107] Construct an augmented matrix based on the first plane equation and the second plane equation;
[0108] Perform Gaussian elimination on the augmented matrix to determine the intersection point position of the first panel and the second panel.
[0109] Determining the panel intersection line of the first panel point cloud and the second panel point cloud according to the intersection line vector and the intersection point position of the first panel and the second panel includes:
[0110] Determine the transformation ratio based on the coordinate component of the intersection line vector along the preset direction and the coordinate value of the intersection point position in the preset direction;
[0111] Determine the panel intersection line of the first panel point cloud and the second panel point cloud based on the transformation ratio and the line segment model.
[0112] Such as Figure 5As shown, an embodiment of the present application provides an electronic device for implementing the radar calibration method in the present application. The device includes a memory, a processor, a bus, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the steps of the above-mentioned radar calibration method are implemented.
[0113] Specifically, the above-mentioned memory and processor can be general memory and processor, and no specific limitation is made here. When the processor runs the computer program stored in the memory, it can execute the above-mentioned radar calibration method.
[0114] Corresponding to the radar calibration method in the present application, an embodiment of the present application also provides a computer-readable storage medium. A computer program is stored on the computer-readable storage medium, and when the computer program is run by a processor, the steps of the above-mentioned radar calibration method are executed.
[0115] Specifically, the storage medium can be a general storage medium, such as a mobile disk, a hard disk, etc. When the computer program on the storage medium is run, it can execute the above-mentioned radar calibration method.
[0116] In the embodiments provided by the present application, it should be understood that the disclosed systems and methods can be implemented in other ways. The system embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there can be other division methods in actual implementation. For another example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some communication interfaces. The indirect coupling or communication connection of the system or unit can be in an electrical, mechanical or other form.
[0117] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units. They can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0118] In addition, each functional unit in the embodiments provided by the present application can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.
[0119] When the above-mentioned functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of this application, in essence, or the part that contributes to the prior art or a part of this technical solution, can be embodied in the form of a software product. This 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.) to execute all or part of the steps of the methods described in various embodiments of this application. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks, or optical discs.
[0120] It should be noted that: similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. In addition, the terms "first", "second", "third", etc. are only used for distinguishing descriptions and cannot be understood as indicating or implying relative importance.
[0121] Finally, it should be noted that: the above-mentioned embodiments are only specific implementation manners of this application, used to illustrate the technical solution of this application, rather than limiting it. The protection scope of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: any person skilled in the art within the technical scope disclosed in this application can still modify the technical solutions recorded in the foregoing embodiments, or can easily think of changes, or perform equivalent replacements on some of the technical features; and these modifications, changes, or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of this application. All should be covered within the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.
Claims
1. A method for radar calibration, characterized in that, the method includes: Obtaining target point clouds of a preset area collected by a target radar of a target vehicle, wherein a preset number of targets are arranged in the preset area, and the targets include a first panel and a second panel at a preset angle; Based on the target point clouds, determining a first panel point cloud and a second panel point cloud corresponding to the first panel and the second panel respectively; Based on the panel intersection line of the first panel point cloud and the second panel point cloud, determining the intersection point of the target and the ground; Based on the intersection point and the reference intersection point corresponding to the intersection point, calibrating the target radar.
2. The method according to claim 1, characterized in that, the determining, based on the target point clouds, of a first panel point cloud and a second panel point cloud corresponding to the first panel and the second panel respectively includes: Performing ground correction on the target point clouds, filtering out some point clouds in the target point clouds to obtain remaining point clouds; Performing clustering processing on the remaining point clouds to obtain the first panel point cloud corresponding to the first panel and the second panel point cloud corresponding to the second panel.
3. The method according to claim 1, characterized in that, the panel intersection line of the first panel point cloud and the second panel point cloud is determined by the following method: Determining an intersection line vector according to a first normal vector of the first panel and a second normal vector of the second panel; Calculating the intersection point of the first panel and the second panel according to a first plane equation of the first panel and a second plane equation of the second panel; Based on the intersection line vector and the intersection point of the first panel and the second panel, determining the panel intersection line of the first panel point cloud and the second panel point cloud.
4. The method according to claim 3, characterized in that, the calculating of the intersection point of the first panel and the second panel according to the first plane equation of the first panel and the second plane equation of the second panel includes: Based on the first plane equation and the second plane equation, constructing an augmented matrix; Performing Gaussian elimination on the augmented matrix to determine the intersection point of the first panel and the second panel.
5. The method according to claim 3, characterized in that, the determining of the panel intersection line of the first panel point cloud and the second panel point cloud according to the intersection line vector and the intersection point of the first panel and the second panel includes: Determining a transformation ratio based on the coordinate component of the intersection line vector along a preset direction and the coordinate value of the intersection point in the preset direction; Based on the transformation ratio and a line segment model, determining the panel intersection line of the first panel point cloud and the second panel point cloud.
6. A device for radar calibration, characterized in that, the device includes: An acquisition module, configured to acquire target point clouds of a preset area collected by a target radar of a target vehicle, wherein a preset number of targets are arranged in the preset area, and the targets include a first panel and a second panel at a preset angle; A first determination module, configured to determine a first panel point cloud and a second panel point cloud corresponding to the first panel and the second panel respectively based on the target point cloud; A second determination module, configured to determine an intersection point of the target and the ground based on an intersection line of the first panel point cloud and the second panel point cloud; A calibration module, configured to calibrate the target radar based on the intersection point and a reference intersection point corresponding to the intersection point; 7. The apparatus according to claim 6, wherein, when determining the first panel point cloud and the second panel point cloud corresponding to the first panel and the second panel respectively based on the target point cloud, the first determination module is configured to: perform ground correction on the target point cloud, filter out some point clouds in the target point cloud, and obtain a remaining point cloud; perform clustering processing on the remaining point cloud to obtain the first panel point cloud corresponding to the first panel and the second panel point cloud corresponding to the second panel.
8. The apparatus according to claim 6, wherein, the second determination module determines the intersection line of the first panel point cloud and the second panel point cloud in the following manner: determine an intersection line vector according to a first normal vector of the first panel and a second normal vector of the second panel; calculate an intersection point of the first panel and the second panel according to a first plane equation of the first panel and a second plane equation of the second panel; determine the intersection line of the first panel point cloud and the second panel point cloud according to the intersection line vector and the intersection point of the first panel and the second panel.
9. An electronic device, wherein, comprises: a processor, a memory, and a bus, the memory stores machine-readable instructions executable by the processor, when the electronic device runs, the processor communicates with the memory through the bus, and when the machine-readable instructions are executed by the processor, the steps of the radar calibration method according to any one of claims 1 to 7 are executed.
10. A computer-readable storage medium, wherein, a computer program is stored on the computer-readable storage medium, and when the computer program is run by a processor, the steps of the radar calibration method according to any one of claims 1 to 7 are executed.