A method, system, device and medium for automatic calibration of multiple radar origin information in freeway traffic scenarios

Abstract

The present application relates to a kind of in expressway traffic scene multi-radar origin information automatic calibration method, system, equipment and medium, its method includes: obtaining the equipment information and trajectory data of radar to be corrected, and judging radar irradiation direction;Obtain and process map data;Determine whether the radar to be corrected road is reference radar;If it is the reference radar of no connection area, adjust the angle of trajectory data after cleaning, and with the road section range in map data matching, according to the number of matched trajectory to determine the best correction angle;If it is the non-reference radar of having connection area, reference the reference radar of relative connection, determine the optimal matching trajectory by cleaning and trajectory matching, and preliminary correction is carried out according to the difference between matching trajectory, according to the matching degree of trajectory data before cleaning in the direction of radar irradiation and road section range secondary correction is carried out.The present application effectively solves the problem in the process of radar calibration by advanced trajectory processing technology and automatic calibration process.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of intelligent transportation technology, and in particular to a method, system, device and medium for automatically calibrating multiple radar origin information in a rapid road traffic scene. BACKGROUND

[0002] Millimeter wave radars have been widely used in modern transportation systems, especially in vehicle detection and monitoring. Due to its stable operation under various weather conditions and the provision of accurate object detection and speed measurement information, millimeter wave radar has become a key sensor in intelligent transportation systems. However, the effectiveness of millimeter wave radar is highly dependent on its installation location, angle, and accurate alignment with the surrounding environment.

[0003] Currently, the installation and calibration of millimeter wave radars are mainly carried out manually. During installation, the initial installation position and angle of the radar are determined using pile numbers, high-precision positioning instruments, and traffic channelization calibration. Then, through a series of conversion processes, the physical position of the radar is mapped onto a digital map to ensure accurate correspondence between radar data and actual geographical position. In addition, a certain number of millimeter wave radar detection trajectories need to be collected and compared with corresponding channelization and visual detection images for matching to achieve accurate calibration of the radar origin.

[0004] However, this manual calibration method has several problems. First, the steps are tedious, especially when working on large projects such as rapid roads and tunnels, as the number of devices involved is large, resulting in high deployment and calibration costs. Second, due to the lack of unified calibration principles, subjective factors have a significant impact on the calibration process, which can lead to inconsistent calibration accuracy and affect the effectiveness of the holographic trajectory. Finally, due to external factors such as vibration and weather, the radar angle and position may shift, so regular inspection and recalibration of the radar origin are required, which incurs high maintenance costs.

[0005] In large-scale holographic projects such as rapid roads, highways, and tunnels, the calibration and adjustment of the installation position and angle of a large number of millimeter wave radar devices become a challenge. If completely relying on manual adjustment, it is easy to cause subjective calibration deviations, waste a large amount of resources, and ultimately affect the detection and connection effect. Therefore, there is an urgent need to improve the millimeter wave radar installation and calibration method. SUMMARY

[0006] (I) Technical problems to be solved

[0007] In view of the above-mentioned shortcomings and deficiencies of the prior art, the present application provides a method, system, device and medium for automatic calibration of multiple radar origin information in an expressway traffic scene, which solves the technical problem that the installation and calibration process of a millimeter wave radar is complicated, high in cost, and easily affected by subjective factors and external environment, resulting in inconsistent calibration accuracy and difficulty in meeting the demand of large-scale holographic projects.

[0008] (II) Technical solutions

[0009] In order to achieve the above-mentioned purposes, the main technical solutions adopted by the present application include:

[0010] In a first aspect, the present application provides a method for automatic calibration of multiple radar origin information in an expressway traffic scene, comprising:

[0011] Obtaining device information and trajectory data of a radar to be corrected at an intersection, and determining a radar irradiation direction;

[0012] Obtaining and processing map data of a road section lane corresponding to the radar to be corrected at the intersection;

[0013] Determining whether the radar to be corrected at the intersection is a reference radar;

[0014] If the reference radar is in a non-connection area, gradually adjusting the angle of the trajectory data after adaptive data cleaning, and matching with the road section range in the map data, determining the best correction angle of the radar origin according to the change in the number of matched trajectories;

[0015] If the non-reference radar has a connection area, taking the reference radar as a reference, determining the optimal matching trajectory through adaptive data cleaning and trajectory matching, and preliminarily correcting the angle and latitude and longitude according to at least one data difference between the optimal matching trajectories, and secondarily correcting the angle and latitude and longitude of the radar origin according to the matching degree of the trajectory data within the radar irradiation direction before adaptive cleaning and the road section range in the map data.

[0016] Optionally, obtaining device information and trajectory data of a radar to be corrected at an intersection, and determining a radar irradiation direction comprises:

[0017] Sending an information acquisition instruction to the radar to be corrected at the intersection to request the device to return data;

[0018] Receiving device information and trajectory data returned by the radar to be corrected at the intersection, and parsing according to a predetermined data format to obtain the required device information and trajectory data;

[0019] Processing and analyzing the obtained trajectory data to extract information related to target motion;

[0020] According to the information related to the target motion, in combination with at least one of the working mode, the scanning mode, and the antenna pointing information of the radar, the irradiation range of the radar is determined.

[0021] Optionally, if the reference radar is without a connection area, the angle of the trajectory data after adaptive data cleaning is adjusted step by step and matched with the range of the road segment in the map data, and the optimal correction angle of the radar origin is determined according to the change in the number of matched trajectories.

[0022] By means of adaptive parameters, the target trajectory to be corrected is selected from the trajectory data of the radar at the intersection to be corrected, which has a trajectory x-coordinate less than a preset threshold, consistent with the driving direction of the road segment, and a trajectory number greater than a set value;

[0023] With a set first coarse adjustment step, the angle of the target trajectory is adjusted step by step within a preset first coarse adjustment angle range, and the trajectory latitude and longitude after angle adjustment are calculated, and the trajectory latitude and longitude after angle adjustment are matched with the range of the road segment in the map data until the number of trajectories in the corrected road segment is reduced, so as to determine the angle after coarse adjustment correction.

[0024] On the basis of the angle after coarse adjustment correction, the angle of the target trajectory is adjusted step by step within a preset first fine adjustment angle range with a set first fine adjustment step, and the trajectory latitude and longitude after angle adjustment are calculated, and the trajectory latitude and longitude after angle adjustment are matched with the range of the road segment in the map data until the number of trajectories in the corrected road segment is reduced again, and finally the optimal correction angle of the radar origin is determined.

[0025] Optionally, if the non-reference radar has a connection area, the reference radar connected therewith is selected as a reference, and the optimal matching trajectory is determined through adaptive data cleaning and trajectory matching, and the angle and latitude and longitude are preliminarily corrected according to at least one data difference between the optimal matching trajectories, and the angle and latitude and longitude of the radar origin are secondarily corrected according to the matching degree of the trajectory data in the irradiation direction of the radar before adaptive cleaning with the range of the road segment in the map data.

[0026] The reference radar connected with the non-reference radar is selected as a reference radar, and the trajectory data of the reference radar and the trajectory data of the non-reference radar to be corrected are adaptively cleaned.

[0027] The trajectory data of the cleaned reference radar and the non-reference radar to be corrected are time-stamped, and the optimal matching trajectory is determined by calculating the speed difference and direction turning angle difference at each time node.

[0028] The angle and latitude and longitude of the radar origin of the non-reference radar to be corrected are preliminarily corrected according to at least one data difference between the optimal matching trajectories.

[0029] The trajectory data of the non-reference radar to be corrected in the direction of radar irradiation before adaptive cleaning is screened and matched with the road section range of the map data, and the angle and the latitude and longitude of the radar origin of the non-reference radar to be corrected are secondarily corrected according to the matching degree.

[0030] Optionally, the adaptive data cleaning on the trajectory data of the reference radar and the trajectory data of the non-reference radar to be corrected comprises:

[0031] For the trajectory data of the reference radar, the adaptive data cleaning is as follows:

[0032] The reference trajectory whose trajectory x coordinate is greater than a first connection comparison value obtained by the maximum value of the trajectory x coordinate minus the connection distance minus a preset buffer value is screened out, so as to ensure that the selected trajectory data is in the connection range;

[0033] According to the road section direction, the matched trajectory data is selected from the reference trajectory, and the interference trajectory with inconsistent direction is excluded;

[0034] The reference trajectory whose number of trajectories is greater than a preset number threshold is screened out;

[0035] The trajectory segment with a direction turning angle greater than a preset angle threshold in the reference trajectory is identified, and the trajectory data meeting the preset floating feature requirement is screened out by calculating the direction turning angle of adjacent points;

[0036] The difference between the head and tail x coordinates of the reference trajectory is calculated, and the trajectory data with a difference greater than a preset buffer value is retained;

[0037] For the trajectory data of the non-reference radar to be corrected, the adaptive data cleaning is as follows:

[0038] The target trajectory whose trajectory x coordinate is less than a second connection comparison value obtained by the minimum value of the x coordinate plus the connection distance plus a preset buffer value is screened out, so as to ensure that the selected trajectory data is in the connection range with the reference radar;

[0039] According to the road section direction, the matched trajectory data is selected from the target trajectory, and the interference trajectory with inconsistent direction is excluded;

[0040] The target trajectory whose number of trajectories is greater than a preset number threshold is screened out;

[0041] The difference between the head and tail x coordinates of the target trajectory is calculated, and the trajectory data with a difference greater than a preset buffer value is retained.

[0042] Optionally, the trajectory data of the reference radar and the non-reference radar to be corrected after cleaning is time stamped, and the optimal matching trajectory is determined by calculating the speed difference and the direction turning angle difference at each time node, comprising:

[0043] The cleaned reference trajectory and the target trajectory are time stamped;

[0044] For each time node after alignment, the time node and the N nodes after the time node are selected to calculate the speed of the reference trajectory and the target trajectory, and the speed difference between the reference trajectory and the target trajectory is calculated according to the speed of the reference trajectory and the target trajectory;

[0045] For each time node after alignment, the time node and the N nodes after the time node are selected to calculate the direction angle of the reference trajectory and the target trajectory, and the direction angle difference between the reference trajectory and the target trajectory is calculated according to the direction angle of the reference trajectory and the target trajectory;

[0046] Read the trajectory segment with a direction angle difference greater than 1 degree in the reference trajectory and the target trajectory, and take the minimum value of the speed difference in the read trajectory segment as the matching coefficient of the two to-be-matched targets;

[0047] According to the minimum matching coefficient principle, find the trajectory segment with the minimum matching coefficient from the read trajectory segment as the optimal matching trajectory.

[0048] Optionally, the preliminary correction of the angle and the latitude and longitude of the radar origin of the to-be-corrected non-reference radar according to at least one data difference between the optimal matching trajectories comprises:

[0049] Compare the reference trajectory and the target trajectory with the optimal matching trajectory respectively, and screen out the complete optimal matching reference trajectory and the complete optimal matching target trajectory containing the optimal matching trajectory segment;

[0050] Calculate the slope according to the head and tail coordinates of the complete optimal matching reference trajectory and the complete optimal matching target trajectory, calculate the arctangent value of the slope, and obtain the heading angle through conversion;

[0051] Preliminary angle correction is performed based on the median of the heading angle difference value of the complete optimal matching reference trajectory and the complete optimal matching target trajectory;

[0052] In the optimal matching trajectory, the latitude and longitude difference between each point is calculated, and from the latitude and longitude difference between each point, a plurality of groups of latitude and longitude differences with difference values within a set range are screened out to calculate the latitude and longitude difference average value;

[0053] The latitude and longitude correction value is used to perform preliminary latitude and longitude correction on the origin latitude and longitude of the non-reference radar;

[0054] Determine whether the angle after preliminary angle correction exceeds the correction angle threshold;

[0055] If the correction angle threshold is exceeded, repeat the preliminary angle correction and the preliminary latitude and longitude correction until the angle correction value is less than the correction angle threshold.

[0056] Optionally, the trajectory data of the non-reference radar to be corrected in the radar irradiation direction before adaptive cleaning is screened and matched with the road section range of the map data, and the angle and the longitude and latitude of the radar origin of the non-reference radar to be corrected are secondarily corrected according to the matching degree, including:

[0057] From the trajectory data of the non-reference radar to be corrected in the radar irradiation direction, trajectory data with an x-coordinate less than a preset correction judgment value is selected.

[0058] The trajectory data less than the preset correction judgment value screened is matched with the road section range of the map data, and whether the matching degree of the trajectory data and the road section range reaches a preset standard is judged.

[0059] If the matching degree reaches the preset standard, the angle is gradually adjusted in a preset second coarse adjustment angle range with a set second coarse adjustment step, and the trajectory longitude and latitude after the angle adjustment is calculated, the trajectory longitude and latitude after the angle adjustment is matched with the road section range in the map data, until the number of trajectories in the corrected road section becomes small, and the angle after coarse adjustment is determined.

[0060] On the basis of the angle after coarse adjustment, the angle is gradually adjusted in a preset second fine adjustment angle range with a set second fine adjustment step, and the trajectory longitude and latitude after the angle adjustment is calculated, the trajectory longitude and latitude after the angle adjustment is matched with the road section range in the map data, until the number of trajectories in the corrected road section becomes small again, and the angle after fine adjustment is determined.

[0061] In the optimal matching trajectory, the longitude and latitude differences between points are calculated, and from the longitude and latitude differences between points, a plurality of groups of longitude and latitude differences with differences in a set range are selected to calculate the average value of the longitude and latitude differences.

[0062] The longitude and latitude correction values are used for secondarily correcting the longitude and latitude of the origin of the non-reference radar.

[0063] In a second aspect, an embodiment of the present application provides a system for automatically calibrating multiple radar origin information in a fast road traffic scene, including:

[0064] An irradiation direction judgment module is configured to acquire device information and trajectory data of a radar at a road intersection to be corrected, and judge the radar irradiation direction.

[0065] A map data acquisition module is configured to acquire and process map data of a road section lane corresponding to the radar at the road intersection to be corrected.

[0066] A reference radar judgment module is configured to judge whether the radar at the road intersection to be corrected is a reference radar.

[0067] A reference radar correction module is used to gradually adjust the angle of the trajectory data after adaptive data cleaning if the reference radar is in a non-connection area, and match the range of the road section in the map data, and determine the best correction angle of the radar origin according to the number of matched trajectories.

[0068] A non-reference module correction module is used to take the reference of the connected reference radar if the non-reference radar is in a connection area, determine the optimal matching trajectory through adaptive data cleaning and trajectory matching, and preliminarily correct the angle and latitude and longitude according to at least one data difference between the optimal matching trajectories, and secondarily correct the angle and latitude and longitude of the radar origin according to the matching degree of the trajectory data in the direction of the radar irradiation and the range of the road section in the map data before adaptive cleaning.

[0069] In a third aspect, an embodiment of the present application provides a junction traffic scene radar and vision real-time queuing trajectory connection device, comprising: at least one database; and a memory in communication connection with the at least one database; wherein the memory stores instructions executable by the at least one database, and the instructions are executed by the at least one database to enable the at least one database to execute the method of automatic calibration of multiple radar origin information in the expressway traffic scene as described above.

[0070] In a fourth aspect, an embodiment of the present application provides a computer readable medium having computer executable instructions stored thereon, and the executable instructions are executed by a processor to implement the method of automatic calibration of multiple radar origin information in the expressway traffic scene as described above.

[0071] (III) Beneficial Effects

[0072] The beneficial effects of the present application are: the present application proposes a brand-new automatic calibration method aiming at the problems of insufficient precision and complex operation in the radar calibration process. The present application is mainly applied to the expressway traffic scene, and realizes the accurate calibration of the radar origin information through advanced technologies such as trajectory clustering and trajectory data distribution fitting. This innovation not only unifies the debugging standard of the radar origin, but also standardizes the entire calibration process, thereby greatly improving the accuracy and efficiency of calibration.

[0073] Specifically, the present application first acquires the device information and trajectory data of the intersection radar to be corrected, and processes in combination with high-precision map data. In the calibration process, the reference radar without connection area and the non-reference radar with connection area are respectively adopted with targeted calibration strategies. The trajectory data is cleaned through adaptive parameters to ensure that the selected trajectory has high confidence and representativeness. Then, the map matching and trajectory matching technologies are used to adjust the origin angle and / or latitude and longitude of the radar at least once until the optimal calibration effect is achieved.

[0074] It is worth mentioning that the automatic calibration method also has the function of periodic triggering, which can provide continuous automatic operation and maintenance support for holographic digital road projects. This means that once the radar equipment deviates or performance declines, it can be discovered and automatically calibrated in time, thereby ensuring the accuracy and stability of radar data. This feature is of great significance for the long-term operation and maintenance of large-scale holographic projects.

[0075] Therefore, the present application effectively solves the technical problems in the radar calibration process through advanced trajectory processing technology and automatic calibration process, and provides strong technical support for the radar application in expressway traffic scenarios. BRIEF DESCRIPTION OF DRAWINGS

[0076] Figure 1 A flowchart of a method for automatic calibration of multiple radar origin information in an expressway traffic scenario according to an embodiment of the present application;

[0077] Figure 2 A specific flowchart of step S1 of a method for automatic calibration of multiple radar origin information in an expressway traffic scenario according to an embodiment of the present application;

[0078] Figure 3 A specific flowchart of step S4 of a method for automatic calibration of multiple radar origin information in an expressway traffic scenario according to an embodiment of the present application;

[0079] Figure 4 A specific flowchart of step S5 of a method for automatic calibration of multiple radar origin information in an expressway traffic scenario according to an embodiment of the present application;

[0080] Figure 5 A specific flowchart of step S51 of a method for automatic calibration of multiple radar origin information in an expressway traffic scenario according to an embodiment of the present application;

[0081] Figure 6 A specific flowchart of step S52 of a method for automatic calibration of multiple radar origin information in an expressway traffic scenario according to an embodiment of the present application;

[0082] Figure 7 A specific flowchart of step S53 of a method for automatic calibration of multiple radar origin information in an expressway traffic scenario according to an embodiment of the present application;

[0083] Figure 8 A specific flowchart of step S54 of a method for automatic calibration of multiple radar origin information in an expressway traffic scenario according to an embodiment of the present application;

[0084] Figure 9A whole flowchart of a method for automatic calibration of multiple radar origin information in a fast road traffic scene is proposed for an embodiment of the present application. DETAILED DESCRIPTION

[0085] In order to better explain the present application, so as to be understood, the present application is described in detail by specific embodiments in combination with the drawings.

[0086] As shown in the drawings, Figure 1 A method for automatic calibration of multiple radar origin information in a fast road traffic scene is proposed in an embodiment of the present application, which comprises: obtaining the device information and trajectory data of the radar to be corrected at the intersection, and judging the radar irradiation direction; obtaining and processing the map data of the road section lane corresponding to the radar to be corrected at the intersection; judging whether the radar to be corrected at the intersection is a reference radar; if it is a reference radar without a connection area, gradually adjusting the angle of the trajectory data after adaptive data cleaning, and matching with the road section range in the map data, determining the best correction angle of the radar origin according to the change in the number of matched trajectories; if it is a non-reference radar with a connection area, taking the reference radar connected as a reference, determining the optimal matching trajectory through adaptive data cleaning and trajectory matching, and preliminarily correcting the angle and latitude and longitude according to at least one data difference between the optimal matching trajectories, and secondarily correcting the angle and latitude and longitude of the radar origin according to the matching degree of the trajectory data in the radar irradiation direction before adaptive cleaning with the road section range in the map data.

[0087] The present application proposes a brand-new automatic calibration method aiming at the problems of insufficient precision and complex operation in the radar calibration process. The present application is mainly applied to the fast road traffic scene, and realizes the accurate calibration of the radar origin information through advanced technologies such as trajectory clustering and trajectory data distribution fitting. This innovation not only unifies the debugging standard of the radar origin, but also standardizes the whole calibration process, thereby greatly improving the accuracy and efficiency of calibration.

[0088] Specifically, the present application first obtains the device information and trajectory data of the radar to be corrected at the intersection, and processes in combination with high-precision map data. In the calibration process, for the reference radar without a connection area and the non-reference radar with a connection area, targeted calibration strategies are respectively adopted. The trajectory data is cleaned through adaptive parameters to ensure that the selected trajectory has high confidence and representativeness. Then, the map matching and trajectory matching technologies are used to adjust the origin angle and / or latitude and longitude of the radar at least once until the optimal calibration effect is achieved.

[0089] It is worth mentioning that the automatic calibration method also has the function of periodic triggering, which can provide continuous automatic operation and maintenance support for holographic digital road projects. This means that once the radar equipment deviates or performance declines, it can be discovered and automatically calibrated in time to ensure the accuracy and stability of radar data. This feature is of great significance for the long-term operation and maintenance of large-scale holographic projects.

[0090] Therefore, the present application effectively solves the technical problems in the radar calibration process through advanced trajectory processing technology and automatic calibration process, and provides strong technical support for the radar application in expressway traffic scenarios.

[0091] In order to better understand the above technical solutions, the exemplary embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present application are shown in the drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided to enable a clearer, more thorough understanding of the present application and to fully convey the scope of the present application to those skilled in the art.

[0092] Specifically, the embodiment of the present application provides a method for automatic calibration of multiple radar origin information in expressway traffic scenarios, which comprises:

[0093] S1, obtaining the device information and trajectory data of the radar to be corrected, and judging the radar irradiation direction.

[0094] Further, as shown in Figure 2 , step S1 comprises:

[0095] S11, sending an information acquisition instruction to the radar to be corrected to request the device to return data.

[0096] S12, receiving the device information and trajectory data returned by the radar to be corrected, and parsing according to the predetermined data format to obtain the required device information and trajectory data.

[0097] Once the data returned by the device is received, it is parsed according to the predetermined data format. These data include the information of the device itself (such as the state of the device, the model of the radar, the serial number, the installation date, the time of the last maintenance or calibration, etc.) and the trajectory data detected by the radar (such as the position of the target object, the moving path, etc.).

[0098] S13, processing and analyzing the obtained trajectory data to extract information related to target motion.

[0099] The purpose of this step is to extract information directly related to target motion from a large amount of raw data, such as the moving trajectory of the target, the speed change, etc.

[0100] S14, determine the illumination range of the radar based on the information related to the target motion, in combination with at least one of the radar's operating mode, scanning method, and antenna pointing information.

[0101] Based on the information extracted in the previous step, the radar's illumination range is determined by combining its operating mode (such as continuous wave, pulsed wave, etc.), scanning method (such as mechanical scanning, electronic scanning, etc.), and antenna pointing information (if available). Additionally, if the radar is equipped with a directional antenna, the antenna pointing information can be directly used to determine the illumination direction, allowing for more accurate tracking and positioning of the target.

[0102] S2, obtain and process map data of the road segment lane corresponding to the radar at the intersection to be corrected.

[0103] Retrieve the map data related to the road segment lane corresponding to the specific intersection radar from the map database or related services. Process the obtained map data. The processing process includes data cleaning to remove errors or redundant information, data formatting to meet the set data processing standards, and data enhancement such as adding lane markers to improve the availability and richness of the data.

[0104] S3, determine whether the radar at the intersection to be corrected is a reference radar.

[0105] In the radar system, the reference radar usually refers to those fixed-position, stable-performance, and accurately calibrated radar devices, whose data is used as a reference for the calibration and correction of other radar devices. Therefore, accurately determining whether an intersection radar is a reference radar is crucial to ensuring the accuracy and reliability of the entire radar system.

[0106] The judgment process first relies on the obtained device information. By comparing the device information with the list of known reference radar devices, possible reference radar candidates can be preliminarily screened out.

[0107] Next, use the radar's coordinate information for further verification. These data can be obtained through GPS or other positioning techniques. Compare these coordinates with the coordinates of known reference radars to check whether they are located at the predetermined reference position.

[0108] At the same time, it is considered that in the expressway traffic scene, the radar without connection area has no reference radar connected to it, so it can only define itself as a reference radar, so it does not judge the non-reference radar without connection area.

[0109] When the radar with the connection area is to be calibrated, the reference radar connected with the radar is selected as the benchmark radar. The non-benchmark radar algorithm includes the high-precision map comparison algorithm and the trajectory matching algorithm, and the calibration accuracy is higher than that of the benchmark radar calibration algorithm. Therefore, the radar with the connection area is generally determined as the non-benchmark radar. Therefore, the benchmark radar without the connection area and the non-benchmark radar with the connection area are judged and corrected next.

[0110] S4, if the benchmark radar without the connection area, the angle of the trajectory data after the adaptive data cleaning is adjusted gradually, and is matched with the road section range in the map data, and the best correction angle of the radar origin is determined according to the number of matched trajectories.

[0111] Further, as shown in Figure 3 , step S4 includes:

[0112] S41, by the adaptive parameter method, the target trajectory to be corrected is selected from the trajectory data of the radar at the intersection to be corrected, which has the x coordinate of the trajectory less than the preset threshold, consistent with the driving direction of the road section and the number of trajectories greater than the set value.

[0113] S42, in the preset first coarse adjustment angle range, the target trajectory angle is adjusted gradually with the set first coarse adjustment step, and the trajectory latitude and longitude after the angle adjustment is calculated. The trajectory latitude and longitude after the angle adjustment is matched with the road section range in the map data, until the number of trajectories in the corrected road section is reduced, so as to determine the angle after the coarse adjustment correction.

[0114] S43, on the basis of the angle after the coarse adjustment correction, the target trajectory angle is adjusted gradually in the preset first fine adjustment angle range with the set first fine adjustment step, and the trajectory latitude and longitude after the angle adjustment is calculated. The trajectory latitude and longitude after the angle adjustment is matched with the road section range in the map data, until the number of trajectories in the corrected road section is reduced again, and finally the best correction angle of the radar origin is determined.

[0115] In an embodiment, the cleaning and angle correction process of the target trajectory data is as follows.

[0116] Firstly, the target trajectory data is cleaned. This process is based on the adaptive parameter method, and the main purpose is to intercept the trajectory with high confidence, corresponding road section driving direction and long appearing time. The specific steps include: selecting the trajectory data with the x coordinate of the trajectory less than 200, which represents high confidence; then, according to the driving direction of the road section, the target trajectory data corresponding thereto is further selected; finally, the trajectory with the number of target trajectories greater than 50 is selected from these trajectories to ensure that the selected trajectory has sufficient representativeness and stability.

[0117] Next, angle correction is performed. This process aims to determine the optimal correction angle by gradually adjusting the trajectory angle and matching it with the road segment range on the map until the number of corrected trajectories is reduced. Angle correction is divided into two stages: coarse adjustment and fine adjustment.

[0118] In the coarse adjustment phase, the angle of the target trajectory is gradually adjusted within a range of -10 to 10, with a step size of 1. After each adjustment, the new latitude and longitude of the trajectory are calculated and matched with the road segment range in the map data. This process continues until the number of trajectories within the corrected road segment decreases. At this point, the coarse adjustment process ends, and the angle before the correction is taken as the angle after the coarse adjustment.

[0119] In the fine-tuning phase, based on the angle corrected by the coarse adjustment, a more precise angle adjustment is performed within a range of -1 to 1, with a step size of 0.1. Similar to the coarse adjustment phase, after each adjustment, the new latitude and longitude of the trajectory are calculated and matched with the road segment range. The fine-tuning process continues until the number of trajectories within the corrected road segment decreases again. At this point, the fine-tuning process ends, and the current angle before correction is determined as the optimal correction angle for the radar origin.

[0120] S5. If there is a non-reference radar with a connecting area, the reference radar with which it is connected is used as a reference. The optimal matching trajectory is determined through adaptive data cleaning and trajectory matching. The angle and latitude and longitude are initially corrected based on at least one data difference between the optimal matching trajectories. The angle and latitude and longitude of the radar origin are then corrected a second time based on the matching degree between the trajectory data within the radar illumination direction before adaptive cleaning and the road segment range of the map data.

[0121] Furthermore, such as Figure 4 As shown, step S5 includes:

[0122] S51. Select a reference radar that is connected to the non-reference radar as the reference radar, and perform adaptive data cleaning on the trajectory data of the reference radar and the trajectory data of the non-reference radar to be corrected.

[0123] Furthermore, such as Figure 5 As shown, step S51 includes:

[0124] For the trajectory data referenced by the radar, the following adaptive data cleaning is performed:

[0125] S511. Filter out reference trajectories whose x-coordinate is greater than the first connection comparison value obtained by subtracting the connection distance from the maximum x-coordinate of the trajectory and subtracting the preset buffer value, to ensure that the selected trajectory data is within the connection range.

[0126] S512. Based on the direction of the road segment, select matching trajectory data from the reference trajectory and eliminate interfering trajectories with inconsistent directions.

[0127] S513, screening out reference trajectories with a number of trajectories greater than a preset number threshold.

[0128] S514, identifying trajectory segments with a direction turning angle greater than a preset angle threshold in the reference trajectories, and calculating the direction turning angle using adjacent point angles to screen out trajectory data meeting preset floating feature requirements;

[0129] S515, calculating the difference between the head and tail x coordinates of the reference trajectories, and retaining trajectory data with a difference value exceeding a preset buffer value.

[0130] In addition, for the trajectory data of the non-reference radar to be corrected, the following adaptive data cleaning is performed:

[0131] S516, screening out target trajectories with an x coordinate less than a second connection comparison value obtained by adding the minimum x coordinate value, the connection distance, and a preset buffer value, to ensure that the selected trajectory data is within the connection range with the reference radar.

[0132] S517, selecting matching trajectory data from the target trajectories according to the road segment direction, and excluding interfering trajectories with inconsistent directions.

[0133] S518, screening out target trajectories with a number of trajectories greater than a preset number threshold.

[0134] S519, calculating the difference between the head and tail x coordinates of the target trajectories, and retaining trajectory data with a difference value exceeding a preset buffer value.

[0135] In an embodiment, the process of adaptive cleaning of the trajectory data of the reference radar and the trajectory data of the non-reference radar to be corrected will be described in detail:

[0136] First, the trajectory data of the reference radar is cleaned. This process aims to intercept reference trajectories meeting specific conditions based on adaptive parameters, including connection range, corresponding road segment driving direction, long appearance time, obvious floating, and large moving distance. The specific steps are as follows: first, screen out trajectory data with an x coordinate greater than (x coordinate maximum value - connection distance - 20) to ensure that the selected trajectory is within the appropriate connection range; then, according to the driving direction of the road segment, further select the target trajectory data corresponding thereto; then, from these trajectories, select trajectories with a target number of trajectories greater than 20 to reflect their long appearance time; in addition, select trajectories with a direction turning angle greater than 1 degree in the target trajectory. In this step, the direction turning angle is calculated using the previous 5 points to ensure that the trajectory has obvious floating; finally, select trajectories with a target head and tail x coordinate difference exceeding 20 to reflect their large moving distance.

[0137] Next, the trajectory data of the non-reference radar to be corrected is cleaned. Similar to the cleaning of the reference trajectory data, this process is also based on adaptive parameters, but the focus is on intercepting the trajectory to be corrected that meets the connection range, the corresponding road segment driving direction, the long appearance time, and the large moving distance. The specific steps include: first, the trajectory data whose x-coordinate is less than (x-coordinate minimum value + connection distance + 20) is selected to determine that it is within the appropriate connection range; second, the target trajectory data of the corresponding driving direction is selected according to the road segment direction; third, the trajectory whose target trajectory number is greater than 20 is selected from these trajectories to reflect its persistence in appearance time; and finally, the trajectory whose target first and last x-coordinate difference exceeds 20 is selected to ensure that it has a large moving distance.

[0138] S52, time stamp alignment is performed on the cleaned trajectory data of the reference radar and the non-reference radar to be corrected, and the optimal matching trajectory is determined by calculating the speed difference and the direction angle difference at each time node.

[0139] Further, as shown in Figure 6 , step S52 includes:

[0140] S521, time stamp alignment is performed on the cleaned reference trajectory and the target trajectory.

[0141] S522, for each time node after alignment, the speed of the reference trajectory and the target trajectory is calculated by selecting the time node and the N nodes after the time node, and the speed difference between the reference trajectory and the target trajectory is calculated according to the speed of the reference trajectory and the target trajectory.

[0142] S523, for each time node after alignment, the direction angle of the reference trajectory and the target trajectory is calculated by selecting the time node and the N nodes after the time node, and the direction angle difference between the reference trajectory and the target trajectory is calculated according to the direction angle of the reference trajectory and the target trajectory.

[0143] S524, the trajectory segment with a direction angle difference greater than 1 degree is read in the reference trajectory and the target trajectory, and the minimum value of the speed difference in the read trajectory segment is taken as the matching coefficient of the two to-be-matched targets.

[0144] S525, according to the minimum matching coefficient principle, the trajectory segment with the minimum matching coefficient is found from the read trajectory segment as the optimal matching trajectory.

[0145] In another specific embodiment, the process of target trajectory matching will be described in detail, which aims to align the time stamp of the reference trajectory and the trajectory to be corrected, and calculate the speed and vector angle at each time node, so as to select the reference trajectory with the smallest speed vector difference from the trajectory to be corrected in the same time period as the optimal matching trajectory.

[0146] Specifically, first, a pre-step of target trajectory matching is performed, i.e., the reference trajectory data is time-stamped aligned with the to-be-corrected trajectory. This is to ensure that in the subsequent calculation, the data points of the two trajectories are corresponding, so that accurate comparison and analysis can be performed.

[0147] Next, for each time node after alignment, two key calculations are performed. One is to select the time node and the 5 points after the time node to calculate the speed of the reference trajectory and the target trajectory. The speed is calculated based on the distance and time difference between each point and the 5th point after it. In this way, the speed value of each time node can be obtained, and the absolute value of the speed difference between the reference trajectory and the target trajectory can be calculated.

[0148] Two is also for each time node, select the node and the 5th point after it to calculate the direction angle of the reference trajectory and the target trajectory. The calculation of the direction angle considers the position relationship between each point and the 5th point before and after it. After calculating the direction angle, we can get the absolute value of the direction angle difference between the reference trajectory and the target trajectory.

[0149] After the above calculations are completed, the data is further processed, the trajectory segment with a direction angle difference greater than 1 degree is read, and the minimum average speed difference value is taken as the matching coefficient of the two targets; according to the minimum matching coefficient principle, the target matching pair is found.

[0150] S53, according to at least one data difference between the optimal matching trajectories, the angle and the latitude and longitude of the radar origin of the to-be-corrected non-reference radar are preliminarily corrected.

[0151] Further, as shown in Figure 7 , step S53 includes:

[0152] S531, compare the reference trajectory and the target trajectory with the optimal matching trajectory respectively, and select the complete optimal matching reference trajectory and the complete optimal matching target trajectory containing the optimal matching trajectory segment.

[0153] S532, calculate the slope according to the head and tail coordinates of the complete optimal matching reference trajectory and the complete optimal matching target trajectory, calculate the arctangent value of the slope, and convert it to get the heading angle.

[0154] Wherein, the heading angle is calculated by the slope, the arctangent value of the slope is calculated, and an angle in radians is obtained. Then, the angle is converted from radians to degrees (180 / π), and the heading angle is obtained. The conversion formula is: heading angle=(180 / π) x atan(slope).

[0155] S533, performing a preliminary angle correction based on the median of the heading angle difference values between the complete optimal matching reference trajectory and the complete optimal matching target trajectory.

[0156] S534, in the optimal matching trajectory, calculating the latitude and longitude difference between each point, and from the latitude and longitude difference between each point, selecting several groups of latitude and longitude difference within the set range to calculate the latitude and longitude difference average value.

[0157] S535, using the latitude and longitude correction value to perform a preliminary latitude and longitude correction on the origin latitude and longitude of the non-reference radar.

[0158] S536, judging whether the angle after preliminary angle correction exceeds the correction angle threshold.

[0159] S537, if it exceeds the correction angle threshold, repeating the preliminary angle correction and preliminary latitude and longitude correction until the angle correction value is less than the correction angle threshold.

[0160] In an embodiment, first, the angle correction step is performed. For each target, the slope and heading angle are calculated based on its first and last coordinates within the matching range. Here, the heading angle represents the direction of the target's movement. Then, the correction angle is calculated using the difference in heading angle between the two sets of radar trajectories, specifically by taking the median of these difference values as the correction value. The purpose of this step is to correct any possible angle deviation and make the trajectory data more accurate.

[0161] Next, the latitude and longitude correction is performed. Between the matching pairs, the latitude and longitude difference between each point is calculated. To obtain more reliable correction data, several groups of data with similar latitude and longitude differences are selected, and their latitude and longitude difference average value is calculated. This average value is then used for latitude and longitude correction to further improve the geographical position accuracy of the trajectory.

[0162] After completing the above two steps of correction, it is checked whether the angle correction value exceeds 0.1 degrees. If it does, it means that the correction has not yet reached the required accuracy, so the steps of origin angle correction and origin latitude and longitude correction need to be repeated. By continuously iterating these two correction processes until the angle correction value is less than 0.1 degrees, it is ensured that the trajectory of the target is accurately calibrated in terms of angle and position. This iterative correction method helps to improve the accuracy and reliability of trajectory correction.

[0163] S54, selecting the trajectory data of the non-reference radar to be corrected in the direction of radar illumination before adaptive cleaning, and matching with the road segment range of the map data, and performing secondary correction on the angle and latitude and longitude of the radar origin of the non-reference radar to be corrected according to the matching degree.

[0164] Further, as shown in Figure 8 , step S54 includes:

[0165] S541、from the trajectory data of the non-reference radar to be corrected which is in the radar irradiation direction, select the trajectory data with x coordinate less than the preset correction judgment value.

[0166] S542、match the selected trajectory data with x coordinate less than the preset correction judgment value with the road section range of the map data, and judge whether the matching degree of the trajectory data and the road section range reaches the preset standard.

[0167] S543、if the matching degree reaches the preset standard, then gradually adjust the angle in the preset second coarse adjustment angle range with the set second coarse adjustment step, calculate the adjusted trajectory latitude and longitude after the angle adjustment, match the adjusted trajectory latitude and longitude after the angle adjustment with the road section range in the map data, until the number of trajectories in the corrected road section becomes smaller, and determine the angle after coarse adjustment.

[0168] S544、on the basis of the angle after coarse adjustment, then gradually adjust the angle in the preset second fine adjustment angle range with the set second fine adjustment step, calculate the adjusted trajectory latitude and longitude after the angle adjustment, match the adjusted trajectory latitude and longitude after the angle adjustment with the road section range in the map data, until the number of trajectories in the corrected road section becomes smaller again, and determine the angle after fine adjustment.

[0169] S545、in the optimal matching trajectory, calculate the latitude and longitude difference between each point, and from the latitude and longitude difference between each point, select the latitude and longitude difference with several groups of difference values within the set range to calculate the average latitude and longitude difference.

[0170] S546、use the latitude and longitude correction value to perform secondary latitude and longitude correction on the original point latitude and longitude of the non-reference radar.

[0171] In a specific embodiment, first, the trajectory data is selected and matched with the range. From the target trajectory data before radar cleaning, select those trajectory data which is in the irradiation direction and whose x coordinate is less than 100. Then, match the selected trajectory data with the road section range. If more than 90% of the data falls within the road section range, it means that the data matching degree is high, and the next correction operation can be continued; otherwise, the process is directly ended, and the pairing failure and correction failure information is prompted.

[0172] After confirming the data matching success, the secondary correction is performed. First, the angle coarse adjustment stage, adjust the angle from -5 to 5 with 1 as the step, and calculate the trajectory latitude and longitude after each adjustment. After each adjustment, match it with the road section range, and observe whether the number of trajectories in the road section becomes smaller. When the number of trajectories begins to decrease, it means that the current angle is no longer the best correction angle, so the coarse adjustment process is ended, and the angle before correction is taken as the angle after correction.

[0173] Next is the angle fine-tuning stage, with a step size of 0.1, the angle is fine-tuned in a smaller range (-1 to 1). Similarly, the latitude and longitude of the trajectory are calculated after each adjustment and matched with the road segment range until the critical point that makes the number of trajectories within the road segment smaller is found. After fine-tuning, the current angle before correction is taken as the final corrected angle.

[0174] Finally, the origin latitude and longitude are corrected. The latitude and longitude differences of each point between the matched pairs are calculated, and several groups of data with similar latitude and longitude differences are selected, and their average values are calculated. The average value is used to correct the origin latitude and longitude to improve the accuracy of the trajectory data.

[0175] After completing the secondary correction, check whether all radars have been calibrated, if there are still uncalibrated radars, continue to calibrate the next connected radar. Through such a loop iteration process, it is ensured that each radar can be accurately calibrated and corrected.

[0176] In addition, the embodiment of the present application provides a system for automatically calibrating origin information of multiple radars in an expressway traffic scene, comprising:

[0177] An irradiation direction judgment module is configured to acquire device information and trajectory data of the radar to be corrected at the intersection, and judge the irradiation direction of the radar;

[0178] A map data acquisition module is configured to acquire and process map data of a road segment lane corresponding to the radar to be corrected at the intersection;

[0179] A reference radar judgment module is configured to judge whether the radar to be corrected at the intersection is a reference radar;

[0180] A reference radar correction module is configured to, if the reference radar has no connection area, gradually adjust the angle of the trajectory data after adaptive data cleaning, and match the road segment range in the map data, and determine the optimal correction angle of the origin of the radar according to the change in the number of matched trajectories;

[0181] A non-reference radar correction module is configured to, if the non-reference radar has a connection area, take the reference radar connected therewith as a reference, determine the optimal matched trajectory through adaptive data cleaning and trajectory matching, and preliminarily correct the angle and latitude and longitude according to at least one data difference between the optimal matched trajectories, and secondarily correct the angle and latitude and longitude of the origin of the radar according to the matching degree of the trajectory data in the irradiation direction of the radar before adaptive cleaning and the road segment range in the map data.

[0182] Furthermore, the embodiment of the present application provides a device for connecting real-time queuing trajectories of radar and vision in an intersection traffic scene, comprising: at least one database; and a memory in communication connection with the at least one database; wherein the memory stores instructions executable by the at least one database, and the instructions are executed by the at least one database to enable the at least one database to perform the method for automatically calibrating multiple radar origin information in a freeway traffic scene as described above.

[0183] When these instructions are executed by the database, they enable the database to perform a series of specific operations, which are exactly the method for automatically calibrating multiple radar origin information in a freeway traffic scene proposed by the present application. This method aims to improve the accuracy and consistency of radar trajectory data by automatically calibrating the origin information of multiple radars, thereby optimizing the monitoring and management of intersection traffic.

[0184] In short, the device provided by the embodiment of the present application combines the functions of database and memory, and realizes accurate calibration of multiple radar origin information in complex traffic scenarios by executing specific instructions, thereby bringing convenience and benefits to traffic monitoring and management.

[0185] Furthermore, the embodiment of the present application provides a computer readable medium having computer executable instructions stored thereon, and the executable instructions are executed by a processor to implement the method for automatically calibrating multiple radar origin information in a freeway traffic scene as described above.

[0186] When these instructions are executed by the processor, they can trigger and implement a specific method, i.e. the method for automatically calibrating multiple radar origin information in a freeway traffic scene described by the present application. This method accurately calibrates the origin information of multiple radars in a freeway traffic scene in an automated manner, thereby ensuring that the trajectory data generated by each radar is optimal in terms of accuracy and consistency. This not only improves the overall performance of the traffic monitoring system, but also provides a more reliable data basis for subsequent traffic flow analysis, congestion prediction and other advanced applications.

[0187] Therefore, this computer readable medium and the instructions stored thereon constitute an indispensable part of the embodiment of the present application, and provide an efficient and convenient solution for calibrating the origin information of a multiple radar system in a freeway traffic scene.

[0188] In summary, the embodiment of the present application provides a method, system, device and medium for automatically calibrating multiple radar origin information in a freeway traffic scene, as Figure 9As shown, the present application cleans and filters the collected trajectory data comprehensively and accurately through the adaptive parameter trajectory data cleaning technology. This step considers multiple dimensions such as the confidence of trajectory data, the length of appearance time, the moving distance, and the consistency of direction, thereby ensuring the reliability and accuracy of the retained data. Secondly, the present application adopts an automatic correction angle technology based on trajectory data and high-precision maps. By gradually adjusting the latitude and longitude of the trajectory and accurately matching with the high-precision map, the automatic and high-precision correction of the radar origin angle is realized. In the calibration process, the present application also uses an advanced trajectory matching algorithm. This algorithm can accurately align and match the trajectory data of the reference radar with the trajectory data of the radar to be corrected. By calculating the speed vector and vector angle difference of each time node, the reference trajectory with the smallest speed vector difference can be accurately found out, ensuring the best matching of the two sets of trajectory data, and the angle calibration based on the high-precision map. In addition, the present application also adopts a method for origin calibration based on the position difference between the reference radar trajectory and the matching trajectory in the radar to be calibrated. By calculating the slope and heading angle of the first and last coordinates in the matching range, and using the median of the heading angle difference of the two sets of radar trajectories for angle correction, combined with the filtering and average calculation of the latitude and longitude difference, the accurate correction of the latitude and longitude is realized.

[0189] The innovation of the present application lies in multiple aspects: by fusing trajectory data clustering and fitting technology, and with the assistance of high-precision maps, the present application has achieved breakthroughs in precision and application range, providing more innovative solutions for complex traffic environments such as expressways. Its high precision and complex scene adaptability benefit from the perfect combination of trajectory data fitting and high-precision maps, enabling multi-radar automatic calibration to perform well under complex conditions.

[0190] At the same time, the automatic calibration method of the present application is extremely convenient to design, can be automatically triggered regularly, significantly reduces the frequency and workload of manual intervention, and improves the overall efficiency of system maintenance. Compared with the existing technology which relies on the periodic appearance of external vehicles, the present application undoubtedly has more advantages.

[0191] Finally, it is worth mentioning that the present application is not only suitable for single-radar systems, but also widely applicable to multi-radar systems, realizing the unification of radar origin calibration standards and the standardization of calibration processes, and showing strong versatility and expansion potential.

[0192] Due to the system / device described in the above embodiments of the present application, the system / device used for implementing the method of the above embodiments of the present application, based on the method described in the above embodiments of the present application, those skilled in the art can understand the specific structure and modification of the system / device, and thus it is not repeated here. Any system / device used in the method of the above embodiments of the present application belongs to the scope of protection of the present application.

[0193] Those skilled in the art will appreciate that embodiments of the present application can be readily used as a method, a system or a computer program product. Accordingly, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) embodying computer readable program code thereon for use by or in connection with an instruction execution system. For the purposes of this description, a computer-usable or computer readable storage medium can be any apparatus that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.

[0194] The present application is described in reference to the flowchart illustrations and / or block diagrams of the methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions of the flowchart and / or block diagrams.

[0195] It should be noted that any references made herein to elements or integers of the specification embodied within the enclosed drawings are intended to be inclusive of common variants of these elements that are discernible to one skilled in the art and are not intended to be limited to the elements presented in the enclosed drawings. It is to be understood that the terms "including", "comprising", "consisting" and variations thereof do not preclude the addition of further integers to those which are recited. It is to be understood that the term "a" or "an", as used herein, does not exclude pluralities and zero of the named item. The present application can be implemented by means of both hardware and software, and any combination of them. In a claim numerating several apparatus, several of these can be embodied by one and the same item of hardware. The terms first, second, third etc. are used merely as labels, and are not intended to impose numerical requirements on their objects. These terms are to be interpreted accordingly.

[0196] Furthermore, it is to be understood that the description of the present application is made with reference to the accompanying drawings, and that the description and drawings are intended to be illustrative of the present application and are not intended to limit the scope of the application. It is to be understood that the description and drawings are not to be taken in a limiting sense. It is to be understood that the description of the present application is made with reference to the accompanying drawings, and that the description and drawings are intended to be illustrative of the present application and are not intended to limit the scope of the application. It is to be understood that the description and drawings are not to be taken in a limiting sense. It is to be understood that the terms "comprise", "comprising", "comprises", "include", "including", and "includes" are to be construed as incorporating by reference the phrase "by not by way of limitation" as if that phrase were written herein. Furthermore, it is to be understood that the description of the present application is made with reference to the accompanying drawings, and that the description and drawings are intended to be illustrative of the present application and are not intended to limit the scope of the application. It is to be understood that the description and drawings are not to be taken in a limiting sense. It is to be understood that the terms "comprise", "comprising", "comprises", "include", "including", and "includes" are to be construed as incorporating by reference the phrase "by not by way of limitation" as if that phrase were written herein.

[0197] Although preferred embodiments of the application have been described herein, it will be apparent to those skilled in the art that variations in these embodiments can be made and that it is intended that the application can be practiced otherwise than as specifically described herein. Accordingly, this application includes all modifications encompassed within the scope of the claims.

[0198] It will be apparent to those skilled in the art that various modifications and variations can be made to the present application without departing from the spirit or scope of the application. Thus, it is intended that the present application cover modifications and variations of this application provided they come within the scope of the appended claims and their equivalents.

Claims

1. A method for automatic calibration of multi-radar origin information in expressway traffic scenarios, characterized in that, include: Acquire the equipment information and trajectory data of the radar at the intersection to be corrected, and determine the radar illumination direction; Acquire and process map data of the road segment lanes corresponding to the intersection radar to be corrected; Determine whether the radar at the intersection to be corrected is the reference radar; If it is a base radar without connecting areas, the angle of the trajectory data after adaptive data cleaning is gradually adjusted and matched with the road segment range in the map data. The optimal correction angle of the radar origin is determined based on the change in the number of matched trajectories. If there is a non-reference radar with a connecting area, the reference radar with which it is connected is used as a reference. The optimal matching trajectory is determined through adaptive data cleaning and trajectory matching. The angle and latitude and longitude are initially corrected based on at least one data difference between the optimal matching trajectories. The angle and latitude and longitude of the radar origin are then corrected a second time based on the matching degree between the trajectory data within the radar illumination direction before adaptive cleaning and the road segment range of the map data.

2. The method for automatic calibration of multi-radar origin information in a fast road traffic scenario as described in claim 1, characterized in that, Acquire equipment information and trajectory data of the radar at the intersection to be corrected, and determine the radar illumination direction, including: Send an information acquisition command to the radar at the intersection to be corrected in order to request the device to return data; Receive device information and trajectory data returned by the radar at the intersection to be corrected, and parse them according to a predetermined data format to obtain the required device information and trajectory data; The acquired trajectory data is processed and analyzed to extract information related to the target's motion; Based on information related to the target's movement, and combined with at least one of the following: radar operating mode, scanning method, and antenna pointing information, the radar's illumination range is determined.

3. The method for automatic calibration of multi-radar origin information in a fast road traffic scenario as described in claim 1, characterized in that, If the base radar has no connecting zones, the angle of the trajectory data after adaptive data cleaning is gradually adjusted and matched with the road segment range in the map data. The optimal correction angle of the radar origin is determined based on the change in the number of matched trajectories, including: By using adaptive parameters, target trajectories to be corrected are selected from the trajectory data of the radar at the intersection to be corrected, whose x-coordinate is less than a preset threshold, whose driving direction is consistent with the road segment, and whose number of trajectories is greater than a set value. With a set first coarse adjustment step size, the target trajectory angle is gradually adjusted within a preset first coarse adjustment angle range, and the latitude and longitude of the trajectory after angle adjustment are calculated. The latitude and longitude of the trajectory after angle adjustment are matched with the road segment range in the map data until the number of trajectories in the corrected road segment is reduced, thereby determining the angle after coarse adjustment. Based on the angle after coarse adjustment, the target trajectory angle is gradually adjusted within the preset first fine adjustment angle range with a set first fine adjustment step size, and the latitude and longitude of the trajectory after angle adjustment are calculated. The latitude and longitude of the trajectory after angle adjustment are matched with the road segment range in the map data until the number of trajectories in the corrected road segment is reduced again, and finally the optimal correction angle of the radar origin is determined.

4. The method for automatic calibration of multi-radar origin information in a fast road traffic scenario as described in any one of claims 1-3, characterized in that, If there are non-reference radars with connecting areas, the connecting reference radars are used as a reference. Through adaptive data cleaning and trajectory matching, the optimal matching trajectory is determined. Based on at least one data difference between the optimal matching trajectories, the angle and latitude / longitude are initially corrected. Furthermore, based on the matching degree between the trajectory data within the radar illumination direction before adaptive cleaning and the road segment range in the map data, the angle and latitude / longitude of the radar origin are further corrected, including: A reference radar that is connected to a non-reference radar is selected as the reference radar, and adaptive data cleaning is performed on the trajectory data of the reference radar and the trajectory data of the non-reference radar to be corrected. The trajectory data of the cleaned reference radar and the non-reference radar to be corrected are time-stamped and aligned. The optimal matching trajectory is determined by calculating the velocity difference and directional angle difference at each time node. Based on at least one data difference between the optimal matching trajectories, the angle and latitude and longitude of the radar origin of the non-reference radar to be corrected are initially corrected; The trajectory data of the non-reference radar to be corrected, which are within the radar illumination direction before adaptive cleaning, are filtered and matched with the road segment range of the map data. Based on the matching degree, the angle and latitude and longitude of the radar origin of the non-reference radar to be corrected are then corrected.

5. The method for automatic calibration of multi-radar origin information in a fast road traffic scenario as described in claim 4, characterized in that, Adaptive data cleaning of the trajectory data from the reference radar and the trajectory data from the non-reference radar to be corrected includes: For the trajectory data referenced by the radar, the following adaptive data cleaning is performed: Filter out reference trajectories whose x-coordinate is greater than the first connection comparison value obtained by subtracting the connection distance from the maximum x-coordinate of the trajectory and the preset buffer value, to ensure that the selected trajectory data is within the connection range; Based on the direction of the road segment, select matching trajectory data from the reference trajectory and eliminate interfering trajectories with inconsistent directions; Filter out reference trajectories whose number of trajectories exceeds a preset threshold; Identify trajectory segments in the reference trajectory whose directional turning angle is greater than a preset angle threshold, and use the angles of adjacent points to calculate the directional turning angle to filter out trajectory data that meet the preset floating feature requirements; Calculate the difference between the first and last x-coordinates of the reference trajectory, and retain trajectory data whose difference exceeds the preset buffer value; For the trajectory data of the non-reference radar to be corrected, the following adaptive data cleaning is performed: Select target trajectories whose x-coordinate is less than the second connection comparison value obtained by adding the minimum x-coordinate value, connection distance, and preset buffer value, to ensure that the selected trajectory data is within the range of connection with the reference radar. Based on the direction of the road segment, select matching trajectory data from the target trajectory and eliminate interfering trajectories with inconsistent directions; Filter out target trajectories whose number of trajectories exceeds a preset threshold; Calculate the difference between the first and last x-coordinates of the target trajectory, and retain trajectory data whose difference exceeds a preset buffer value.

6. The method for automatic calibration of multi-radar origin information in a fast road traffic scenario as described in claim 5, characterized in that, The trajectory data of the cleaned reference radar and the non-reference radar to be corrected are timestamped and aligned. By calculating the velocity difference and azimuth angle difference at each time point, the optimal matching trajectory is determined, including: The cleaned reference trajectory and the target trajectory are timestamped together. For each aligned time node, select that time node and the N nodes after that time node to calculate the velocity of the reference trajectory and the target trajectory, and calculate the velocity difference between the reference trajectory and the target trajectory based on the velocity of the reference trajectory and the target trajectory. For each aligned time node, select that time node and the N nodes after that time node to calculate the directional rotation angles of the reference trajectory and the target trajectory, and calculate the difference in directional rotation angles between the reference trajectory and the target trajectory based on the directional rotation angles of the reference trajectory and the target trajectory; Read trajectory segments with a directional turning angle difference greater than 1 degree from the reference trajectory and the target trajectory, and use the minimum velocity difference in the read trajectory segments as the matching coefficient between the two targets to be matched; Based on the principle of minimum matching coefficient, the trajectory segment with the minimum matching coefficient is found from the read trajectory segments and is taken as the optimal matching trajectory.

7. The method for automatic calibration of multi-radar origin information in a fast road traffic scenario as described in claim 6, characterized in that, Based on at least one data difference between the optimal matching trajectories, the initial correction of the angle and latitude / longitude of the radar origin of the non-reference radar to be corrected includes: The reference trajectory and the target trajectory are compared with the optimal matching trajectory respectively, and the complete optimal matching reference trajectory and the complete optimal matching target trajectory containing the optimal matching trajectory segment are selected. The slope is calculated based on the first and last coordinates of the complete optimal matching reference trajectory and the complete optimal matching target trajectory. The heading angle is obtained by calculating the arctangent of the slope and converting it. Preliminary angle correction is performed based on the median of the heading angle difference between the complete optimal matching reference trajectory and the complete optimal matching target trajectory; In the optimal matching trajectory, the latitude and longitude difference between each point is calculated. From the latitude and longitude differences between each point, several sets of latitude and longitude differences within a set range are selected to calculate the average latitude and longitude difference. The latitude and longitude of the origin of the non-reference radar is initially corrected using latitude and longitude correction values; Determine whether the angle after initial angle correction exceeds the correction angle threshold; If the correction angle threshold is exceeded, the initial angle correction and initial latitude and longitude correction will be repeated until the angle correction value is less than the correction angle threshold.

8. The method for automatic calibration of multi-radar origin information in a fast road traffic scenario as described in claim 7, characterized in that, The trajectory data of the non-reference radar to be corrected, which are within the radar illumination direction before adaptive cleaning, are filtered and matched with the road segment range of the map data. Based on the matching degree, a secondary correction is performed on the angle and latitude / longitude of the radar origin of the non-reference radar to be corrected, including: From the trajectory data of the non-reference radar to be corrected within the radar illumination direction, select trajectory data whose x-coordinate is less than the preset correction judgment value; The selected trajectory data that is less than the preset correction judgment value is matched with the road segment range of the map data to determine whether the matching degree between the trajectory data and the road segment range meets the preset standard. If the matching degree reaches the preset standard, the angle is gradually adjusted within the preset second coarse adjustment angle range using the set second coarse adjustment step size, and the latitude and longitude of the trajectory after angle adjustment are calculated. The latitude and longitude of the trajectory after angle adjustment are matched with the road segment range in the map data until the number of trajectories in the corrected road segment becomes smaller, and the angle after coarse adjustment is determined. Based on the coarsely adjusted angle, the angle is gradually adjusted within the preset second fine adjustment angle range using the set second fine adjustment step size, and the latitude and longitude of the trajectory after the angle adjustment are calculated. The latitude and longitude of the trajectory after the angle adjustment are matched with the road segment range in the map data until the number of trajectories in the corrected road segment becomes smaller again, and the finely adjusted angle is determined. In the optimal matching trajectory, the latitude and longitude difference between each point is calculated. From the latitude and longitude differences between each point, several sets of latitude and longitude differences within a set range are selected to calculate the average latitude and longitude difference. The latitude and longitude of the origin of the non-reference radar is corrected using latitude and longitude correction values.

9. A system for automatic calibration of multi-radar origin information in expressway traffic scenarios, characterized in that, include: The illumination direction determination module is used to acquire the equipment information and trajectory data of the radar at the intersection to be corrected, and to determine the radar illumination direction. The map data acquisition module is used to acquire and process map data of the road segment lanes corresponding to the intersection radar to be corrected. The reference radar determination module is used to determine whether the radar at the intersection to be corrected is a reference radar. The baseline radar correction module is used to gradually adjust the angle of the trajectory data after adaptive data cleaning if the baseline radar has no connecting area, and match it with the road segment range in the map data. The optimal correction angle of the radar origin is determined based on the change in the number of matched trajectories. The non-reference module correction module is used to determine the optimal matching trajectory by using the reference radar as a reference when there is a non-reference radar with a connecting area. Based on the data difference of at least one data between the optimal matching trajectories, the angle and latitude and longitude are initially corrected. Furthermore, based on the matching degree between the trajectory data within the radar illumination direction before adaptive cleaning and the road segment range of the map data, the angle and latitude and longitude of the radar origin are corrected a second time.

10. A connection device for real-time queuing trajectory monitoring via radar in an intersection traffic scenario, characterized in that, include: At least one database; The system also includes a memory that is communicatively connected to at least one database; wherein the memory stores instructions that can be executed by at least one database to enable at least one database to perform the method for automatic calibration of multi-radar origin information in a fast road traffic scenario as described in any one of claims 1-8.

11. A computer-readable medium having computer-executable instructions stored thereon, characterized in that, When the executable instructions are executed by the processor, they implement the method for automatic calibration of multi-radar origin information in a fast road traffic scenario as described in any one of claims 1-8.

Images