A LiDAR Calibration Method as a Truth System Evaluation

By using calibration vehicles and markers on a flat ground, the problems of expensive equipment and difficult operation in the prior art are solved, and a low threshold and efficient lidar calibration method is realized.

CN119758308APending Publication Date: 2025-04-04HUAYU AUTOMOTIVE SYST
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411842652.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-12-13
Publication Date
2025-04-04

AI Technical Summary

Technical Problem

The existing lidar calibration method requires high-priced equipment such as four-wheel positioning devices, special calibration plates and calibration workshops, which makes it difficult to operate in the early research and development.

Method used

A low-threshold lidar calibration method is provided. By selecting a calibration site on a relatively flat ground, using calibration vehicles and markers for calibration, the lidar calibration coordinate system is determined, and the coordinate position of the lidar is obtained by collecting and processing data.

Benefits of technology

It reduces the equipment and conditions required for lidar calibration, improves the convenience and efficiency of operation, and is suitable for scenarios with limited resources in the early research and development.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119758308A_ABST
    Figure CN119758308A_ABST
Patent Text Reader

Abstract

The present invention relates to a laser radar calibration method as a true value system evaluation, comprising: step S1, selecting a calibration site, parking a calibration vehicle in the calibration site, and installing a laser radar to be calibrated on the top of the calibration vehicle, and determining the laser radar calibration coordinate system; step S2, selecting marking points in the calibration site, and recording the coordinate position of each marking point in the laser radar calibration coordinate system; step S3, selecting markers, and placing the markers according to the coordinate positions of the marking points in the laser radar calibration coordinate system, so that the contours of all markers appear in the visualization interface of the laser radar to be calibrated; step S4, when the calibration vehicle is in a stationary and stable state, the laser radar to be calibrated collects data within a certain time period; step S5, processing the collected data, obtaining the coordinate position of the laser radar to be calibrated installed on the calibration vehicle, and completing the calibration of the laser radar to be calibrated.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of intelligent driving technology, and more specifically to a laser radar calibration method as a true value system evaluation. Background Art

[0002] With the popularization of intelligent driving, the demand for sensors in Advanced Driving Assistance System (ADAS) is increasing. As an all-weather, small-sized, and stable sensor, millimeter-wave radar has become an indispensable component in the field of intelligent driving technology.

[0003] When developing products, millimeter-wave radar suppliers need to evaluate the accuracy of their target detection. The commonly used method is to use the GPS positioning system for evaluation, such as the RT equipment developed by OXTS. However, this type of equipment is not only expensive, but also requires joint testing with the target vehicle on which the equipment is installed. Limited to the two-vehicle test scenario, it is impossible to perform multi-target true value evaluation in actual road scenarios.

[0004] As an active detection technology, LiDAR can accurately measure the three-dimensional information of the target. Therefore, LiDAR can be used as a true value system to evaluate millimeter-wave radar, which will help suppliers conduct better tests, obtain objective and accurate results, and assist the R&D team to make millimeter-wave radar more accurate. To evaluate millimeter-wave radar using LiDAR, it is necessary to calibrate the LiDAR first. The existing calibration method usually uses a four-wheel alignment device to first fix the vehicle position, then place calibration plates around it, and adjust the calibration parameters through the host computer. This method has high requirements for equipment and requires a four-wheel alignment device, a special calibration plate, a calibration workshop, etc. It is difficult to operate if these conditions are not met in the early stage of R&D. Summary of the invention

[0005] In order to solve the above problems in the prior art, the present invention provides a lidar calibration method as a true value system evaluation to evaluate the target accuracy of other sensors such as millimeter wave radar.

[0006] The present invention provides a laser radar calibration method as a true value system evaluation, comprising:

[0007] Step S1, selecting a calibration site, parking a calibration vehicle in the calibration site, installing a laser radar to be calibrated on the top of the calibration vehicle, and determining a laser radar calibration coordinate system;

[0008] Step S2, selecting marking points in the calibration site, and recording the coordinate position of each marking point in the laser radar calibration coordinate system;

[0009] Step S3, selecting markers, and placing the markers according to the coordinate positions of the marker points in the laser radar calibration coordinate system, so that the outlines of all the markers appear in the visualization interface of the laser radar to be calibrated;

[0010] Step S4, when the calibrated vehicle is in a stationary and stable state, the laser radar to be calibrated collects data within a certain time period;

[0011] Step S5, processing the data collected in step S4, obtaining the coordinate position of the laser radar to be calibrated installed on the calibration vehicle, and completing the calibration of the laser radar to be calibrated.

[0012] Furthermore, determining the laser radar calibration coordinate system in step S1 includes:

[0013] Step S11, taking the center axis of the vehicle to be calibrated as a reference, measuring a first auxiliary line parallel to the center axis of the vehicle as the X-axis of the laser radar calibration coordinate system;

[0014] Step S12, measuring the rear axle line of the calibration vehicle as a second auxiliary line, which serves as the Y axis of the laser radar calibration coordinate system;

[0015] Step S13: The intersection of the first auxiliary line and the second auxiliary line is the coordinate origin of the laser radar calibration coordinate system, and the coordinate origin is marked in the calibration site.

[0016] Furthermore, both the first auxiliary line and the second auxiliary line are measured using a laser measuring instrument.

[0017] Furthermore, the line connecting the marking points is parallel to the X-axis or Y-axis of the laser radar calibration coordinate system.

[0018] Furthermore, the marker is a carton.

[0019] Furthermore, placing the marker in step S3 includes: aligning a side edge of the carton toward the laser radar to be calibrated, and placing a corner point corresponding to the side edge facing the laser radar to be calibrated on one of the marking points.

[0020] Furthermore, the processing of the data collected in step S4 in step S5 includes: parsing the data collected in step S4 into a PCD point cloud file, and obtaining the vehicle body external parameter parameters of the laser radar to be calibrated by comparing the data with the real coordinates of the marked points.

[0021] The laser radar calibration method of the present invention only requires a relatively flat ground and several markers, has a low threshold for requirements and is easy to operate, and the markers are placed in a rectangular shape, which is more convenient for measuring point coordinates. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 is a flow chart of a laser radar calibration method as a truth value system evaluation according to the present invention.

[0023] Figure 2 It is a schematic diagram of the calibration site layout. DETAILED DESCRIPTION

[0024] The preferred embodiments of the present invention are given below in conjunction with the accompanying drawings and described in detail.

[0025] like Figure 1 As shown, the present invention provides a laser radar calibration method as a true value system evaluation, comprising the following steps:

[0026] Step S1, select a calibration site, park the calibration vehicle in the calibration site, install the laser radar to be calibrated on the top of the calibration vehicle, and determine the laser radar calibration coordinate system. The calibration site is a relatively flat ground.

[0027] Specifically, determining the laser radar calibration coordinate system includes:

[0028] Step S11, taking the center axis of the vehicle to be calibrated as a reference, measure a first auxiliary line parallel to the center axis of the vehicle as the X-axis of the laser radar calibration coordinate system. The distance between the first auxiliary line and the center axis of the vehicle can be determined according to the actual situation, as long as a marker can be placed in the calibration site.

[0029] Step S12, measuring the rear axle line of the calibration vehicle as the second auxiliary line, which is used as the Y axis of the laser radar calibration coordinate system. In the embodiment of the present invention, both the first auxiliary line and the second auxiliary line are measured using a laser measuring instrument.

[0030] Step S13: The intersection of the first auxiliary line and the second auxiliary line is the coordinate origin of the laser radar calibration coordinate system, and the coordinate origin is marked in the calibration site.

[0031] Step S2, select a marking point in the calibration site, and record the coordinate position of each marking point in the laser radar calibration coordinate system. Each marking point is outside the blind area of ​​the laser radar to be calibrated.

[0032] To improve the overall efficiency of calibration, the marking points can be selected on the sides of a rectangle, such as Figure 2As shown in the figure, the marking point is the position of "×" shown in the figure, that is, the line connecting the marking points is parallel to the X-axis or Y-axis of the laser radar calibration coordinate system. Selecting the marking points on the sides of a rectangle can make the markers placed in a rectangular shape, making it easier to measure the coordinates of the marking points: the marking points parallel to the X-axis have the same Y-axis coordinates, so only the X-axis coordinates need to be measured, and the same is true for the marking points parallel to the Y-axis.

[0033] Step S3, select markers, and place the markers according to the coordinate positions of the marker points in the laser radar calibration coordinate system so that the outlines of all markers appear in the visualization interface of the laser radar to be calibrated.

[0034] In order to avoid the problem of laser point cloud reflection and point cloud expansion, it is necessary to ensure that the surface of the marker is diffusely reflective. At the same time, in order to ensure the accuracy of laser radar calibration, in the embodiment of the present invention, a carton with sharp edges and corners is selected as the marker.

[0035] When placing the cartons, it also includes: placing a side edge of the carton towards the laser radar to be calibrated, and placing the corner point corresponding to the side edge facing the laser radar to be calibrated on one of the marking points, so that the laser radar to be calibrated can detect the outline of the marker.

[0036] Step S4, when the calibrated vehicle is in a stationary and stable state, the laser radar to be calibrated collects data within a certain period of time (for example, 10 seconds).

[0037] Before the laser radar collects data, other irrelevant items in the calibration area need to be cleared, leaving only the cardboard box, and the personnel leave the calibration area, return to the calibration vehicle, and close the door. After the person sits in the vehicle, the vehicle will shake to a certain extent, and the laser radar installed on it will also shake with it. It is necessary to wait until the laser radar is completely stable before collecting data.

[0038] Step S5, processing the data collected in step S4, obtaining the coordinate position of the laser radar to be calibrated installed on the calibration vehicle, and completing the calibration of the laser radar to be calibrated.

[0039] Processing the data collected in step S4 includes: parsing the data collected in step S4 into a PCD point cloud file, and obtaining the vehicle body external parameter parameters of the laser radar to be calibrated by comparing it with the real coordinates of the marked points, that is, the coordinate position of the laser radar to be calibrated installed on the calibration vehicle.

[0040] After the lidar calibration is completed, road testing can be carried out to evaluate the target accuracy of the millimeter wave radar, including:

[0041] 1) The laser radar and millimeter wave radar collect data synchronously, and then use the large model algorithm of running laser to identify the tracking target, and finally generate a series of text files containing target information;

[0042] 2) After the laser target is generated, the targets of the laser radar and millimeter-wave radar can be seen in real time by replaying it through the simulator software; or the target information of the two can be exported, and then the evaluation logic can be designed (for example, limiting the area, considering the overlap between the two as a successful match, etc.), and a script can be written to evaluate the two. For example, an evaluation tool can be designed based on the Python language. After the target information file is imported, an evaluation report can be automatically issued, including the target's false detection rate, missed detection rate, classification accuracy, etc., to provide a true value basis for evaluating the target accuracy of the millimeter-wave radar, without the need to manually review the data, greatly improving R&D efficiency.

[0043] The above is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. The above embodiment of the present invention can also be modified in various ways. That is, all simple, equivalent changes and modifications made according to the claims and the description of the present invention fall within the scope of protection of the claims of the present invention. The contents not described in detail in the present invention are all conventional technical contents.

Claims

1. A laser radar calibration method as a truth system evaluation, characterized in that: include: Step S1, selecting a calibration site, parking a calibration vehicle in the calibration site, installing a laser radar to be calibrated on the top of the calibration vehicle, and determining a laser radar calibration coordinate system; Step S2, selecting marking points in the calibration site, and recording the coordinate position of each marking point in the laser radar calibration coordinate system; Step S3, selecting markers, and placing the markers according to the coordinate positions of the marker points in the laser radar calibration coordinate system, so that the outlines of all the markers appear in the visualization interface of the laser radar to be calibrated; Step S4, when the calibrated vehicle is in a stationary and stable state, the laser radar to be calibrated collects data within a certain time period; Step S5, processing the data collected in step S4, obtaining the coordinate position of the laser radar to be calibrated installed on the calibration vehicle, and completing the calibration of the laser radar to be calibrated.

2. The laser radar calibration method as a true value system evaluation according to claim 1, characterized in that: Determining the laser radar calibration coordinate system in step S1 includes: Step S11, taking the center axis of the vehicle to be calibrated as a reference, measuring a first auxiliary line parallel to the center axis of the vehicle as the X-axis of the laser radar calibration coordinate system; Step S12, measuring the rear axle line of the calibration vehicle as a second auxiliary line, which serves as the Y axis of the laser radar calibration coordinate system; Step S13: The intersection of the first auxiliary line and the second auxiliary line is the coordinate origin of the laser radar calibration coordinate system, and the coordinate origin is marked in the calibration site.

3. The laser radar calibration method as a true value system evaluation according to claim 2, characterized in that: The first auxiliary line and the second auxiliary line are both measured using a laser measuring instrument.

4. The laser radar calibration method as a true value system evaluation according to claim 2, characterized in that: The line connecting the marking points is parallel to the X-axis or Y-axis of the laser radar calibration coordinate system.

5. The laser radar calibration method as a true value system evaluation according to claim 1, characterized in that: The marker is a carton.

6. The laser radar calibration method as a true value system evaluation according to claim 5, characterized in that: Placing the marker in step S3 includes: aligning a side edge of the carton toward the laser radar to be calibrated, and placing a corner point corresponding to the side edge facing the laser radar to be calibrated on one of the marking points.

7. The laser radar calibration method as a true value system evaluation according to claim 1, characterized in that: The processing of the data collected in step S4 in step S5 includes: parsing the data collected in step S4 into a PCD point cloud file, and obtaining the vehicle body external parameter parameters of the laser radar to be calibrated by comparing the data with the real coordinates of the marked points.