A method, apparatus, and device for generating tunnel maps based on radar data.

Abstract

This application discloses a method, apparatus, and device for generating tunnel maps based on radar data, comprising the following steps: acquiring moving target point cloud data from each radar within the tunnel, and fitting the point cloud data to obtain lane line coordinates; synchronizing target trajectory data from each radar within the tunnel, and using the trajectory coordinates of moving targets appearing simultaneously in the overlapping area of ​​adjacent radars to estimate the installation position coordinate deviation and orientation angle deviation of adjacent radars; calibrating the lane line coordinates using the radar installation position coordinate deviation and orientation angle deviation to ensure uniformity of lane line coordinates for all radars within the tunnel; and transmitting the latitude and longitude information acquired by radars outside the tunnel to the lane line map within the tunnel to obtain a tunnel map with latitude and longitude information. The method of this application automatically generates lane maps from target data collected by radar, eliminating the need for manual map information collection via satellite systems, and is an efficient and low-cost method for solving the problem of tunnel map generation.

Description

Technical Field

[0001] This application belongs to the field of radar detection technology, specifically relating to a method, apparatus, and device for generating tunnel maps based on radar data. Background Technology

[0002] Normally, road maps are obtained by surveying vehicles using satellite positioning systems to locate and collect lane information throughout the entire route. However, inside tunnels, due to the enclosed environment and lack of satellite signals, lane maps cannot be obtained through positioning systems, thus affecting the acquisition of lane maps within the tunnel. Summary of the Invention

[0003] Purpose of the invention: This application provides a method, apparatus, and device for generating tunnel maps based on radar data, aiming to solve the problem that tunnel maps are not easy to obtain in the prior art.

[0004] Technical Solution: This application provides a method for generating tunnel maps based on radar data, including the following steps:

[0005] Acquire moving target point cloud data from each radar inside the tunnel, and fit the point cloud data to obtain lane line coordinates;

[0006] By synchronizing the target trajectory data of each radar in the tunnel, and using the trajectory coordinates of moving targets that appear simultaneously in the overlapping area of ​​adjacent radars, the installation position coordinate deviation and orientation angle deviation of adjacent radars are estimated.

[0007] The lane line coordinates are calibrated by the radar installation position coordinate deviation and orientation angle deviation to ensure that the lane line coordinates of all radars in the tunnel are consistent.

[0008] The latitude and longitude information obtained by radar outside the tunnel is transmitted to the lane line map inside the tunnel to obtain a tunnel map with latitude and longitude information.

[0009] In some embodiments, the radar satisfies at least one of the following conditions:

[0010] 1) The radar is provided in a total of M groups; of which, there is 1 group of radar outside the tunnel and N groups of radar inside the tunnel, and N+1=M, where N and M are integers greater than 0;

[0011] 2) The installation distance between adjacent radars located in the tunnel is am, and the effective range of the radar located in the tunnel is bm, satisfying: a < b;

[0012] 3) The power overlap area of ​​adjacent radars located in the tunnel is cm, and the radar blind zone of the radar located in the tunnel is dm, satisfying: (cd) > 20.

[0013] In some embodiments, the step of acquiring moving target point cloud data from each radar within the tunnel and fitting lane line coordinates to the point cloud data further includes:

[0014] Clustering is performed on the point cloud data of each radar, and the density center of the point cloud distribution is extracted as the sampling point of the lane center line. The point cloud data contains the position coordinate information of the moving target relative to the radar.

[0015] The lane centerline sampling points of each radar are fitted to obtain the fitted curve of the lane centerline of each radar.

[0016] Based on the number of lanes and lane width, the fitted curve is copied and laterally shifted to obtain the coordinates of all lane lines.

[0017] In some embodiments, the step of estimating the installation position coordinate deviation and orientation angle deviation of adjacent radars by using the target trajectory data of each radar within the synchronous tunnel and the trajectory coordinates of moving targets appearing simultaneously in the overlapping area of ​​adjacent radars further includes:

[0018] Based on the moving distance of the moving target, the trajectory coordinates of moving targets appearing simultaneously in adjacent radar overlap areas are collected and denoted as... and ;in, This represents the trajectory coordinate data of the i-th radar inside the tunnel. Represents the trajectory coordinate data of the (i+1)th radar inside the tunnel, where 1≤i≤N and i is an integer;

[0019] Collected separately and Perform linear fitting to obtain the fitting coefficients. and ;in, This represents the slope of the fitted line for the i-th radar. Let represent the slope of the fitted line for the (i+1)th radar. This represents the intercept of the fitted line for the i-th radar. This represents the intercept of the fitted line for the (i+1)th radar.

[0020] Based on the fitting coefficient and Obtain the installation position coordinate deviation of adjacent radars. and the difference in orientation angle .

[0021] In some embodiments, the fitting coefficient and The acquisition is achieved by constructing a data matrix. and Then there is ,and ;

[0022] Coordinate deviation of installation positions of adjacent radars The formula for obtaining it is: ,and ;

[0023] Orientation angle difference The formula for obtaining it is: ;

[0024] in, and These represent the starting ordinates of the trajectory data.

[0025] In some embodiments, the step of calibrating lane line coordinates by radar installation position coordinate deviation and orientation angle deviation to unify the lane line coordinates of all radars within the tunnel further includes:

[0026] The installation position coordinate deviation obtained by adjacent radars inside the tunnel and the difference in orientation angle Stored in vector , and middle;

[0027] Search using sliding window search algorithm , , The location with the highest data density is used as the estimated coordinate deviation and orientation angle deviation, denoted as follows: , and

[0028] Using the calculated deviation estimate, the lane line coordinates of the (i+1)th radar are calibrated to obtain the calibrated lane line coordinates of the (i+1)th radar. ;

[0029] Repeat the above steps to calibrate the lane line map coordinates of radar group M, obtaining the calibrated lane line map coordinate data, denoted as... .

[0030] In some embodiments, vector The expression is: ,vector The expression is: ,vector The expression is: Where L0 represents the data length.

[0031] In some embodiments, the lane line coordinates of the calibrated (i+1)th radar The expression is: .

[0032] In some embodiments, middle, , .

[0033] In some embodiments, the step of transmitting latitude and longitude information acquired by radar outside the tunnel to lane line map inside the tunnel to obtain a tunnel map with latitude and longitude information further includes:

[0034] Obtain the longitude, latitude, and north angle of the radar outside the tunnel, denoted as... ;

[0035] Based on the WGS84 coordinate system, Convert to geocentric rectangular coordinate system ,Right now ,in This indicates the conversion of latitude and longitude to geocentric rectangular coordinates.

[0036] Using the external radar of the tunnel as a reference, the calibrated lane line coordinates are converted into geocentric rectangular coordinates. ,in, ;

[0037] Convert the map's geocentric rectangular coordinate system to latitude and longitude, i.e. ,in, express inverse transform, It indicates the longitude, latitude, and north angle of the lane line map.

[0038] In some embodiments, this application also provides a tunnel map generation apparatus based on radar data, comprising:

[0039] The acquisition module is configured to acquire moving target point cloud data of each radar in the tunnel and fit the point cloud data to obtain lane line coordinates.

[0040] The estimation module is configured to synchronize the target trajectory data of each radar in the tunnel and estimate the installation position coordinate deviation and orientation angle deviation of adjacent radars by using the trajectory coordinates of moving targets that appear simultaneously in the overlapping area of ​​adjacent radars.

[0041] The calibration module is configured to calibrate the lane line coordinates by the radar installation position coordinate deviation and orientation angle deviation, so as to unify the lane line coordinates of all radars in the tunnel.

[0042] A conversion module is configured to transmit latitude and longitude information acquired by radar outside the tunnel to a lane map inside the tunnel to obtain a tunnel map with latitude and longitude information.

[0043] In some embodiments, this application also provides a tunnel map generation device based on radar data, including a processor and a memory storing computer program instructions; the processor executes the computer program instructions to implement the tunnel map generation method based on radar data.

[0044] Beneficial effects compared to existing technologies: This application provides a tunnel map generation method based on radar data, comprising the following steps: acquiring moving target point cloud data from each radar within the tunnel, and fitting the point cloud data to obtain lane line coordinates; synchronizing target trajectory data from each radar within the tunnel, and using the trajectory coordinates of moving targets appearing simultaneously in the overlapping areas of adjacent radars to estimate the installation position coordinate deviation and orientation angle deviation of adjacent radars; calibrating the lane line coordinates using the radar installation position coordinate deviation and orientation angle deviation to ensure uniformity of lane line coordinates for all radars within the tunnel; and transmitting the latitude and longitude information acquired by radars outside the tunnel to the lane line map within the tunnel to obtain a tunnel map with latitude and longitude information. This method automatically generates lane maps from target data collected by radar, eliminating the need for manual map information collection via satellite systems, and is an efficient and low-cost method for solving the problem of tunnel map generation.

[0045] It is understood that, compared with the prior art, the tunnel map generation device and the tunnel map generation equipment based on radar data provided in this application have all the technical features and beneficial effects of the above-mentioned tunnel map generation method based on radar data, and will not be repeated here. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0047] Figure 1 This is a flowchart of the tunnel map generation method based on radar data provided in the embodiments of this application;

[0048] Figure 2 This is a schematic diagram of tunnel map generation provided in an embodiment of this application;

[0049] Figure 3 This is a schematic diagram of a tunnel map generation device based on radar data provided in an embodiment of this application. Detailed Implementation

[0050] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0051] In the description of this application, it should be noted that the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0052] Please see Figure 1 This paper provides a method for generating tunnel maps based on radar data, including the following steps:

[0053] S1: Acquire point cloud data of moving targets from each radar in the tunnel, and fit the point cloud data to obtain lane line coordinates;

[0054] S2: Synchronize the target trajectory data of each radar in the tunnel, and use the trajectory coordinates of moving targets that appear simultaneously in the overlapping area of ​​adjacent radars to estimate the installation position coordinate deviation and orientation angle deviation of adjacent radars.

[0055] S3: The lane line coordinates are calibrated by the radar installation position coordinate deviation and orientation angle deviation to make the lane line coordinates of all radars in the tunnel uniform;

[0056] S4: Transmit the latitude and longitude information obtained by the radar outside the tunnel to the lane line map inside the tunnel to obtain a tunnel map with latitude and longitude information.

[0057] Understandably, lane line coordinate fitting helps achieve accurate lane line identification and positioning, providing drivers or autonomous driving systems with accurate lane information, thereby improving driving safety and experience. By synchronizing target trajectory data from various radars within the tunnel and estimating the installation position coordinate deviations and orientation angle deviations of adjacent radars, errors in radar installation positions can be corrected, improving the accuracy and consistency of the radar system. Using the estimated radar installation position coordinate deviations and orientation angle deviations, the lane line coordinates of all radars within the tunnel can be unified, helping to eliminate coordinate differences between different radars and improving lane line consistency and accuracy. By transmitting latitude and longitude information acquired by radars outside the tunnel to the lane line map inside the tunnel, integration of the tunnel map with other geographic information systems can be achieved, providing more comprehensive and accurate navigation and positioning services. Therefore, the method of this application automatically generates lane maps from target data collected by radar, eliminating the need for manual map information collection via satellite systems. This improves the accuracy, consistency, and reliability of the radar system within the tunnel, providing a safer and more precise driving experience. It is an efficient and low-cost method for solving the challenge of tunnel map generation.

[0058] In some embodiments, the radars satisfy the following conditions: there are M groups of radars in total; wherein, there is 1 group of radars outside the tunnel and N groups of radars inside the tunnel, and N+1=M, where N and M are integers greater than 0. Figure 2 For example, suppose N radars are deployed inside the tunnel. The radar outside the tunnel is numbered 1, and the remaining radars inside the tunnel are numbered 2 to N. Here, i and i+1 are the numbers of any two adjacent radars among the N radars.

[0059] In some embodiments, the radars satisfy the following conditions: the installation distance between adjacent radars located within the tunnel is am, the effective range of the radars located within the tunnel is bm, and a < b; the effective overlap area of ​​adjacent radars located within the tunnel is cm, and the radar blind zone of the radars located within the tunnel is dm, and (cd) > 20. To achieve map stitching, the radar installation spacing must be less than the radar's effective range, and the effective overlap area of ​​adjacent radars must cover the radar blind zone plus 20m. For example, if the blind zone of a tunnel radar in the outermost lane is 20m, then the overlap area of ​​adjacent radars must be greater than 40m.

[0060] Understandably, the effective range of a radar system refers to the maximum distance at which it can effectively detect and measure targets; it represents the farthest detection range achievable under ideal conditions. In a radar system, the overlapping area of ​​adjacent radars needs to be extended by 20 meters beyond the coverage area of ​​the radar's blind zone. This is to ensure that there are no undetected areas within the radar system's coverage. The coverage area of ​​a radar system is typically determined by factors such as the radar's transmit power, receive sensitivity, antenna gain, and target characteristics. However, radar systems may have blind zones in certain situations—areas where targets cannot be effectively detected. These blind zones may be caused by the radar's antenna directivity, terrain obstruction, multipath effects, or other interference factors. To ensure that there are no undetected areas between adjacent radars, a certain overlap area needs to be added beyond the coverage of the radar's blind zone. This ensures that when a target moves from the coverage area of ​​one radar to the coverage area of ​​another, no detection is missed. Typically, to guarantee the coverage capability of the overlapping area, a certain safety distance is added beyond the coverage of the radar's blind zone. This safety distance is usually 20 meters, but the specific value may vary depending on the actual application and system requirements.

[0061] In some embodiments, the step of acquiring moving target point cloud data from each radar within the tunnel and fitting the point cloud data to obtain lane line coordinates further includes:

[0062] Cluster the point cloud data of each radar and extract the density center of the point cloud distribution as the sampling point of the lane center line. The point cloud data contains the position coordinate information of the moving target relative to the radar.

[0063] The lane centerline sampling points of each radar are fitted to obtain the fitted curve of the lane centerline of each radar.

[0064] Based on the number of lanes and lane width, the fitted curve is copied and laterally shifted to obtain the coordinates of all lane lines.

[0065] Understandably, by clustering the point cloud data of each radar, the density center of the point cloud distribution can be extracted. This helps to determine the approximate location of the lane centerline, providing a reference line for the driver or autonomous driving system. Fitting the lane centerline sampling points of each radar helps to accurately determine the shape and curvature of the lane centerline, providing more accurate lane information to the driver or autonomous driving system. Based on the number of lanes and lane width, copying and laterally shifting the fitted curve can yield the coordinates of all lane lines. This helps to determine the lane line positions on multi-lane roads, providing accurate lane line information to the driver or autonomous driving system, thereby improving driving safety and driving experience.

[0066] In some embodiments, the step of estimating the installation position coordinate deviation and orientation angle deviation of adjacent radars by using the target trajectory data of each radar within the synchronous tunnel and the trajectory coordinates of moving targets appearing simultaneously in the overlapping area of ​​adjacent radars further includes:

[0067] Based on the moving distance of the moving target, the trajectory coordinates of moving targets appearing simultaneously in adjacent radar overlap areas are collected and denoted as... and ;in, This represents the trajectory coordinate data of the i-th radar inside the tunnel. Represents the trajectory coordinate data of the (i+1)th radar inside the tunnel, where 1≤i≤N and i is an integer;

[0068] Collected separately and Perform linear fitting to obtain the fitting coefficients. and ;in, This represents the slope of the fitted line for the i-th radar. Let represent the slope of the fitted line for the (i+1)th radar. This represents the intercept of the fitted line for the i-th radar. This represents the intercept of the fitted line for the (i+1)th radar.

[0069] Based on the fitting coefficient and Obtain the installation position coordinate deviation of adjacent radars. and the difference in orientation angle .

[0070] Understandably, obtaining the installation position coordinate deviations and orientation angle differences between adjacent radars aims to correct the positioning errors of the radar system. By estimating and correcting the positional deviations between radars, the positioning accuracy of the radar system can be improved, ensuring a smooth transition and consistency of targets between different radars. Furthermore, it facilitates the fusion and consistency of radar data. By understanding the positional and orientation differences between radars, data from different radars can be calibrated and aligned, ensuring data consistency and accuracy, thereby improving the reliability and effectiveness of radar data. Therefore, determining the installation position coordinate deviations and orientation angle differences between adjacent radars can improve the accuracy of target tracking and positioning, and also provide consistent environmental perception and decision support, thereby improving the decision-making accuracy and robustness of the driving system.

[0071] In some embodiments, the fitting coefficient and The acquisition is achieved by constructing a data matrix. and Then there is ,and .

[0072] In some embodiments, the installation position coordinate deviation of adjacent radars The formula for obtaining it is: ,and ,in, and These represent the starting ordinates of the trajectory data.

[0073] In some embodiments, the orientation angle difference The formula for obtaining it is: .

[0074] In some embodiments, the step of calibrating lane line coordinates by radar installation position coordinate deviation and orientation angle deviation to unify the lane line coordinates of all radars within the tunnel further includes:

[0075] The L0 group installation position coordinate deviation obtained by adjacent radars inside the tunnel and the difference in orientation angle Stored in vector , and In the middle; where L0 above represents the data length, that is, the number of data points obtained;

[0076] Search using sliding window search algorithm , , The location with the highest data density is used as the estimated coordinate deviation and orientation angle deviation, denoted as follows: , and ;

[0077] Using the calculated deviation estimate, the lane line coordinates of the (i+1)th radar are calibrated to obtain the calibrated lane line coordinates of the (i+1)th radar. ;

[0078] Repeat the above steps to calibrate the lane line map coordinates of radar group M, obtaining the calibrated lane line map coordinate data, denoted as... .

[0079] As we can understand, the Sliding Window Search Algorithm is an algorithm for object detection and localization. Its function is to locate objects in images or videos and determine their bounding boxes or regions. It offers advantages such as flexibility, multi-scale search, object localization, and scalability, making it suitable for object detection tasks of different sizes and shapes. This algorithm works by sliding a fixed-size window across the image or video, and then applying an object detection algorithm or classifier at each window location to determine whether a target exists within the window. The window is typically a rectangular or square region and can be adjusted according to the size and shape of the target.

[0080] In some embodiments, vector The expression is: ,vector The expression is: ,vector The expression is: .

[0081] In some embodiments, the lane line coordinates of the calibrated (i+1)th radar The expression is: .

[0082] In some embodiments, middle, , .

[0083] In some embodiments, the step of transmitting latitude and longitude information acquired by radar outside the tunnel to lane line map inside the tunnel to obtain a tunnel map with latitude and longitude information further includes:

[0084] Obtain the longitude, latitude, and north angle of the radar outside the tunnel, denoted as... ;

[0085] Based on the WGS84 coordinate system, Convert to geocentric rectangular coordinate system ,Right now ,in This indicates the conversion of latitude and longitude to geocentric rectangular coordinates.

[0086] Using the external radar of the tunnel as a reference, the calibrated lane line coordinates are converted into geocentric rectangular coordinates. ,in, ;

[0087] Convert the map's geocentric rectangular coordinate system to latitude and longitude, i.e. ,in, express inverse transform, It indicates the longitude, latitude, and north angle of the lane line map.

[0088] Understandably, the above steps further unify the coordinate system, converting both radar and map data to a geocentric rectangular coordinate system, simplifying data processing and analysis. Coordinate transformations between different data sources can be more easily compared and integrated, improving data consistency and reliability. Furthermore, converting lane line coordinates to latitude and longitude provides accurate geolocation information, crucial for applications such as navigation, route planning, and map display, ensuring the vehicle's precise location and orientation on the map. Simultaneously, converting map data to latitude and longitude allows for integration with other geographic information systems. For example, it enables interaction with GPS data, geodatabases, or online map services, achieving more comprehensive and accurate navigation and positioning services. Additionally, tunnel maps with latitude and longitude information can be more easily shared and collaborated with other systems or platforms, allowing different applications and services to share map data for more efficient data exchange and collaborative work.

[0089] In some embodiments, see Figure 3 This application provides a tunnel map generation device based on radar data, comprising:

[0090] The acquisition module is configured to acquire point cloud data of moving targets from each radar in the tunnel and fit the point cloud data to obtain lane line coordinates.

[0091] The estimation module is configured to synchronize the target trajectory data of each radar in the tunnel and use the trajectory coordinates of moving targets that appear simultaneously in the overlapping area of ​​adjacent radars to estimate the installation position coordinate deviation and orientation angle deviation of adjacent radars.

[0092] The calibration module is configured to calibrate the lane line coordinates by measuring the radar installation position coordinate deviation and orientation angle deviation, so as to unify the lane line coordinates of all radars in the tunnel.

[0093] The conversion module is configured to transmit the latitude and longitude information acquired by the radar outside the tunnel to the lane line map inside the tunnel, thereby obtaining a tunnel map with latitude and longitude information.

[0094] Understandably, the advantages of the acquisition module are that it can acquire point cloud data of moving targets from various radars within the tunnel, providing a basis for fitting lane line coordinates and obtaining accurate lane line coordinates for subsequent processing and applications; the advantages of the estimation module are that it can correct the positioning errors of the radar system and improve the accuracy of target tracking and positioning; the advantages of the calibration module are that it ensures the consistency and accuracy of the calibrated lane line coordinates, providing more reliable lane information to the driver or autonomous driving system; and the advantages of the conversion module are that it enables the tunnel map with latitude and longitude information to provide accurate geolocation and integration with other geographic information systems, achieving more comprehensive and accurate navigation and positioning services.

[0095] It should be noted that the tunnel map generation device based on radar data in this embodiment is a system corresponding to the tunnel map generation method based on radar data described above. The functional modules in the device correspond to the respective steps in the prediction method. The tunnel map generation device based on radar data in this embodiment can be implemented in conjunction with the tunnel map generation method based on radar data described above. Furthermore, the functional modules described above can be fully or partially integrated into a single physical entity, or they can be physically separated. These modules can be implemented entirely in software via processing element calls; they can be fully implemented in hardware; or some modules can be implemented by processing element calls to software, while others are implemented in hardware. Moreover, these modules can be fully or partially integrated together, or implemented independently. The processing element here can be an integrated circuit with signal processing capabilities. During implementation, some or all steps of the above method, or the functional modules mentioned above, can be completed through integrated logic circuits in the hardware of the processor element or through software instructions.

[0096] In some embodiments, the present application provides a tunnel map generation device based on radar data. The device may include a processor and a memory storing computer program instructions. The processor may include a central processing unit or a specific integrated circuit, or may be configured to implement one or more integrated circuits of the present application embodiments.

[0097] Memory may include mass storage for data or instructions. For example, and not limitingly, memory may include hard disk drives, floppy disk drives, flash memory, optical disk drives, magneto-optical disk drives, magnetic tape drives, or Universal Serial Bus drives, or combinations of two or more of these. Where appropriate, memory may include removable or non-removable (or fixed) media. Where appropriate, memory may be internal or external to an electronic device. In some specific embodiments, memory is non-volatile solid-state memory.

[0098] In some specific embodiments, the memory may include read-only memory (ROM), random access memory (RAM), disk storage media devices, optical storage media devices, flash memory devices, and electrical, optical, or other physical / tangible memory storage devices. Therefore, typically, the memory includes one or more tangible computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed, it is operable to perform the tunnel map generation method based on radar data according to this embodiment.

[0099] In some embodiments, the electronic device may further include a communication interface and a bus. The processor, memory, and communication interface are connected via the bus and communicate with each other. The communication interface is mainly used to realize communication between various modules, devices, units, and / or equipment in the embodiments of this application.

[0100] In some embodiments, when a processor executes computer program instructions, it performs the following steps:

[0101] S1: Acquire point cloud data of moving targets from each radar in the tunnel, and fit the point cloud data to obtain lane line coordinates;

[0102] S2: Synchronize the target trajectory data of each radar in the tunnel, and use the trajectory coordinates of moving targets that appear simultaneously in the overlapping area of ​​adjacent radars to estimate the installation position coordinate deviation and orientation angle deviation of adjacent radars.

[0103] S3: The lane line coordinates are calibrated by the radar installation position coordinate deviation and orientation angle deviation to make the lane line coordinates of all radars in the tunnel uniform;

[0104] S4: Transmit the latitude and longitude information obtained by the radar outside the tunnel to the lane line map inside the tunnel to obtain a tunnel map with latitude and longitude information.

[0105] In a specific embodiment, this application is based on a tunnel structure. Since the installation spacing and orientation of radars inside the tunnel are unknown, it is necessary to use the target data of adjacent radars to estimate the coordinate deviation and orientation angle deviation, and use the deviation estimate to compensate for the coordinates of the generated radar lane line map. The tunnel map generation method based on radar data is described in detail below.

[0106] (1) Collect moving target point cloud data from each radar, which contains the target's position coordinates relative to the radar;

[0107] (2) Cluster the point cloud data of each radar and extract the density center of the point cloud distribution as the sampling point of the lane center line;

[0108] (3) Perform third-order polynomial fitting on the lane centerline sampling points of each radar to obtain the fitting curve of the lane centerline of each radar.

[0109] (4) For each radar, based on the known number of lanes and lane width, the fitted lane lines are copied and translated laterally to obtain the coordinates of all lane lines;

[0110] (5) Synchronize target trajectory data from all radars;

[0111] (6) When a target appears on radar i+1, the trajectory coordinate data of radar i+1 and radar i with a length of L are collected simultaneously, and denoted as follows: and Where L=20m, it means that the target has moved a distance of 20m;

[0112] (7) Perform linear fitting on the collected trajectory data respectively to obtain its fitting coefficient. and ;

[0113] (8) Calculate the coordinate deviation of the installation positions of adjacent radars. and the difference in orientation angle ;in, , , ;

[0114] (9) Repeat steps (5) to (8) a total of L0 times, and save the difference calculated in step (8) in a vector. , and middle;

[0115] (10) Search using sliding window search algorithm , , The location with the highest data density is used as the estimated coordinate deviation and orientation angle deviation, denoted as follows: , and ;

[0116] (11) Calculate Set search window ,calculate minimum value and maximum value ;by For step, from arrive Perform a sliding search and count the number of data points that fall into the window each time, denoted as . Find The position of the maximum value is denoted as . The corresponding data location is It will fall into the range The average of the data within is obtained. ;

[0117] (12) Calculate Set search window ,calculate minimum value and maximum value ;by For step, from arrive Perform a sliding search and count the number of data points that fall into the window each time, denoted as . Find The position of the maximum value is denoted as . The corresponding data location is It will fall into the range The average of the data within is obtained. ;

[0118] (13) Calculate Set search window ,calculate minimum value and maximum value ;by For step, from arrive Perform a sliding search and count the number of data points that fall into the window each time, denoted as . Find The position of the maximum value is denoted as . The corresponding data location is It will fall into the range The average of the data within is obtained. ;

[0119] (14) The lane line map of radar i+1 is calibrated using the calculated deviation value. Specifically, the lane line coordinate data matrix of radar i+1 is set as follows: The calibrated lane coordinates are: ;

[0120] (15) Following steps (6) to (14), perform deviation calibration on the lane line map coordinates of all radars to obtain the calibrated lane line map coordinate data, denoted as .in, , Since the origin is taken as the first radar (i.e., the radar located outside the tunnel), the lane coordinates of the first radar do not need to be calibrated. This represents the lane line coordinates of the first radar that does not require calibration, while the others correspond to the lane line coordinates of the i-th radar after calibration.

[0121] (16) According to the latitude and longitude coordinates and the first radar ( Figure 2 The conversion relationship of the leftmost radar (located outside the tunnel) is used as a reference to convert the lane line map coordinates to latitude and longitude, resulting in a lane line map with latitude and longitude. Specifically, the latitude, longitude, and north angle of the first radar outside the tunnel are known, and the longitude, latitude, and north angle are represented as follows: Based on the WGS84 coordinate system, Convert to geocentric rectangular coordinate system ,Right now Using the first radar as a reference, the calibrated lane line coordinates are converted into geocentric rectangular coordinates. ,in, Finally, the map's geocentric rectangular coordinate system is converted into latitude and longitude, i.e. Among them, Lon ’ map Lat represents the longitude of the lane line map. ’ map Indicates the latitude of the lane line map, Ang ’ map This indicates the north corner of the lane line map.

[0122] In summary, the above methods can automatically generate lane maps by collecting target data without relying on satellite systems for manual map information collection. This method is efficient and low-cost. The obtained map information can be further applied in navigation, safety, environmental perception, vehicle positioning, and GIS integration, providing drivers and autonomous driving systems with accurate location information and decision support.

[0123] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0124] The foregoing has provided a detailed description of a method, apparatus, and device for generating tunnel maps based on radar data, and specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the technical solutions and core ideas of this application. 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; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A method for generating tunnel maps based on radar data, characterized in that, Includes the following steps: Acquire moving target point cloud data from each radar inside the tunnel, and fit the point cloud data to obtain lane line coordinates; By synchronizing the target trajectory data of each radar in the tunnel, and using the trajectory coordinates of moving targets that appear simultaneously in the overlapping area of ​​adjacent radars, the installation position coordinate deviation and orientation angle deviation of adjacent radars are estimated. The lane line coordinates are calibrated by the radar installation position coordinate deviation and orientation angle deviation to ensure that the lane line coordinates of all radars in the tunnel are consistent. The latitude and longitude information obtained by radar outside the tunnel is transmitted to the lane line map inside the tunnel to obtain a tunnel map with latitude and longitude information.

2. The tunnel map generation method based on radar data according to claim 1, characterized in that, The radar satisfies at least one of the following conditions: 1) The radar is provided in a total of M groups; of which, there is 1 group of radar outside the tunnel and N groups of radar inside the tunnel, and N+1=M, where N and M are integers greater than 0; 2) The installation distance between adjacent radars located in the tunnel is am, and the effective range of the radar located in the tunnel is bm, satisfying: a < b; 3) The power overlap zone of adjacent radars located in the tunnel is cm, and the radar blind zone of the radar located in the tunnel is dm, satisfying: (cd) > 20.

3. The tunnel map generation method based on radar data according to claim 1, characterized in that, The step of acquiring moving target point cloud data from each radar within the tunnel and fitting lane line coordinates to the point cloud data further includes: Clustering is performed on the point cloud data of each radar, and the density center of the point cloud distribution is extracted as the sampling point of the lane center line. The point cloud data contains the position coordinate information of the moving target relative to the radar. The lane centerline sampling points of each radar are fitted to obtain the fitted curve of the lane centerline of each radar. Based on the number of lanes and lane width, the fitted curve is copied and laterally shifted to obtain the coordinates of all lane lines.

4. The tunnel map generation method based on radar data according to claim 2, characterized in that, The steps of synchronizing target trajectory data from various radars within the tunnel, and estimating the installation position coordinate deviation and orientation angle deviation of adjacent radars using the coordinates of moving targets appearing simultaneously within the overlapping area of ​​adjacent radars, further include: Based on the moving distance of the moving target, the trajectory coordinates of moving targets appearing simultaneously in adjacent radar overlap areas are collected and denoted as... and ;in, This represents the trajectory coordinate data of the i-th radar inside the tunnel. Represents the trajectory coordinate data of the (i+1)th radar inside the tunnel, where 1≤i≤N and i is an integer; Collected separately and Perform linear fitting to obtain the fitting coefficients. and ;in, This represents the slope of the fitted line for the i-th radar. Let represent the slope of the fitted line for the (i+1)th radar. This represents the intercept of the fitted line for the i-th radar. This represents the intercept of the fitted line for the (i+1)th radar. Based on the fitting coefficient and Obtain the installation position coordinate deviation of adjacent radars. and the difference in orientation angle .

5. The tunnel map generation method based on radar data according to claim 4, characterized in that, Fit coefficients and The acquisition is as follows: Constructing a data matrix and Then there is ,and .

6. The tunnel map generation method based on radar data according to claim 4, characterized in that, Coordinate deviation of installation positions of adjacent radars The formula for obtaining it is: ,and ;in, and These represent the starting ordinates of the trajectory data.

7. The tunnel map generation method based on radar data according to claim 4, characterized in that, Orientation angle difference The formula for obtaining it is: .

8. The tunnel map generation method based on radar data according to claim 4, characterized in that, The step of calibrating lane line coordinates by measuring radar installation position coordinate deviations and orientation angle deviations to ensure uniform lane line coordinates across all radars within the tunnel further includes: The installation position coordinate deviation obtained by adjacent radars inside the tunnel and the difference in orientation angle Stored in vector , and In the middle; vector The expression is: ,vector The expression is: ,vector The expression is: Where L0 represents the data length; a sliding window search algorithm is used for searching. , , The location with the highest data density is used as the estimated coordinate deviation and orientation angle deviation, denoted as follows: , and ; Using the calculated deviation estimate, the lane line coordinates of the (i+1)th radar are calibrated. Let the lane line coordinate data matrix of radar (i+1) be... The lane coordinates of the (i+1)th radar after calibration are obtained. ; Repeat the above steps to calibrate the lane line map coordinates of radar group M, obtaining the calibrated lane line map coordinate data, denoted as... .

9. The tunnel map generation method based on radar data according to claim 8, characterized in that, Lane coordinates of the (i+1)th radar after calibration The expression is: .

10. The tunnel map generation method based on radar data according to claim 8, characterized in that, Further satisfy: ， ； in, This represents the lane line coordinates of the first radar that does not require calibration; the others... This corresponds to the lane line coordinates of the i-th radar after calibration.

11. The tunnel map generation method based on radar data according to claim 10, characterized in that, The step of transmitting latitude and longitude information acquired by radar outside the tunnel to the lane line map inside the tunnel to obtain a tunnel map with latitude and longitude information further includes: Obtain the longitude, latitude, and north angle of the radar outside the tunnel, denoted as... ; Based on the WGS84 coordinate system, Convert to geocentric rectangular coordinate system ,Right now ,in This indicates the conversion of latitude and longitude to geocentric rectangular coordinates. Using the external radar of the tunnel as a reference, the calibrated lane line coordinates are converted into geocentric rectangular coordinates. ,in, ; Convert the map's geocentric rectangular coordinate system to latitude and longitude, i.e. ,in, express inverse transform, It indicates the longitude, latitude, and north angle of the lane line map.

12. A tunnel map generation device based on radar data, characterized in that, include: The acquisition module is configured to acquire moving target point cloud data of each radar in the tunnel and fit the point cloud data to obtain lane line coordinates. The estimation module is configured to synchronize the target trajectory data of each radar in the tunnel and estimate the installation position coordinate deviation and orientation angle deviation of adjacent radars by using the trajectory coordinates of moving targets that appear simultaneously in the overlapping area of ​​adjacent radars. The calibration module is configured to calibrate the lane line coordinates by the radar installation position coordinate deviation and orientation angle deviation, so as to unify the lane line coordinates of all radars in the tunnel. A conversion module is configured to transmit latitude and longitude information acquired by radar outside the tunnel to a lane map inside the tunnel to obtain a tunnel map with latitude and longitude information.

13. A tunnel map generation device based on radar data, characterized in that, This includes the processor and the memory that stores computer program instructions; When the processor executes computer program instructions, it implements the tunnel map generation method based on radar data as described in any one of claims 1 to 8.

Images