A road lane line generation method, device, computer equipment and storage medium

Abstract

The present application belongs to the field of intelligent traffic information technology, and relates to a road lane line generation method, device, computer equipment and storage medium. The method generates road lane lines by collecting raw data from millimeter wave radar, processing vehicle trajectory data according to longitudinal segmentation of roads, identifying cross-section road lane lines and identifying continuous road lane lines. The present application effectively solves the disadvantages of the traditional lane channelization drawing method, which is time-consuming and labor-intensive, and has huge labor costs.

Description

Technical Field

[0001] The present application relates to the field of intelligent traffic information technology, and in particular to a road lane line generation method, device, computer equipment and storage medium. Background Art

[0002] Intelligent transportation system is an important field in the construction of modern smart cities. It is an intelligent transportation management system that combines and applies information technologies such as big data and cloud computing technology, sensors and radio frequency identification technology, embedded systems and communication technology to the transportation system, and integrates advanced transportation information technology to build an intelligent transportation management system. The construction goal of the intelligent transportation system is to grasp the traffic information of the urban road network in real time, accurately obtain the trajectory dynamics of road vehicles, and quickly make predictions and warnings for traffic problems on the road, and even make autonomous decisions to reduce the accident rate and achieve scientific control of road traffic. Road dynamic traffic flow information is the data basis for building an intelligent transportation system. Holographic traffic perception of various elements and participants of the road network is the inevitable direction of the development of the intelligent transportation system. Therefore, accurate and comprehensive holographic traffic perception of roads provides macro and micro data and decision-making support for traffic situation analysis, traffic control decisions, and traffic information release, which is an urgent need for the construction of an intelligent transportation system.

[0003] There is an auxiliary lane line drawing method, which is to manually measure data and conduct on-site evaluation of the actual radar installation scene, judge, match and measure the lane attributes of each lane in the road with the radar installation position as the starting point, write it into the system through manual input, and the host computer software draws it in the radar data coordinate system.

[0004] However, the applicant found that due to the long-term external force on the radar, there is an angle offset between the normal of the radar millimeter-wave transmitting surface and the actual lane direction, causing the vehicle trajectory to deviate from the lane area division configured when the radar was first installed. Equipment maintenance personnel are still required to perform regular measurements and adjustments. This shows that the traditional auxiliary lane line drawing method has the problems of high labor costs and low drawing accuracy. Summary of the invention

[0005] The purpose of the embodiments of the present application is to propose a road lane line generation method, device, computer equipment and storage medium to solve the problems of high labor cost and low drawing accuracy in traditional auxiliary lane line drawing methods.

[0006] In order to solve the above technical problems, the present application provides a method for generating road lane lines, which adopts the following technical solutions:

[0007] Receive the original vehicle trajectory data collected by the radar collection device;

[0008] Performing a denoising operation on the original vehicle trajectory data to obtain denoised vehicle trajectory data;

[0009] Performing a segment division operation on the denoised vehicle trajectory data to obtain N segment division trajectory data, where N is an integer greater than or equal to 1;

[0010] Performing Gaussian function fitting operations on the N segment division trajectory data respectively to obtain N fitting result parameters;

[0011] Calculating the variance values ​​of the N fitting result parameters, and performing a linear function fitting operation on the variance values ​​to obtain a measurement error variance;

[0012] Extracting the peak value of each lane in the measurement error variance according to a multi-peak Gaussian fitting algorithm, and performing Gaussian distribution fitting on each of the peak values ​​to obtain a cross-sectional trajectory distribution curve, wherein the trajectory distribution curve is composed of Gaussian functions of each peak of the multi-peak Gaussian function;

[0013] Acquire the cross-section lane line position corresponding to the cross-section trajectory distribution curve in the cross-section lane line distribution sample library according to the similarity algorithm;

[0014] Generate a target road lane line according to the cross-section lane line position.

[0015] Furthermore, the step of performing a denoising operation on the original vehicle trajectory data to obtain denoised vehicle trajectory data specifically includes the following steps:

[0016] The original vehicle trajectory data is corrected according to a data correction algorithm to obtain the denoised vehicle trajectory data, wherein the data correction algorithm is expressed as:

[0017]

[0018]

[0019] in, and They represent the lateral distance and radial distance of the target detected by the radar, respectively; x and y represent the corrected lateral distance and radial distance, respectively; and β represents the angle to be rotated.

[0020] Furthermore, after the step of calculating the variance values ​​of the N fitting result parameters and performing a linear function fitting operation on the variance values ​​to obtain the measurement error variance, and before the step of extracting the peak values ​​of each lane in the measurement error variance according to a multi-peak Gaussian fitting algorithm and performing Gaussian distribution fitting on each peak value to obtain a cross-sectional trajectory distribution curve, the following steps are also included:

[0021] The measurement error variance is smoothed according to a Kalman filter algorithm.

[0022] Further, the multi-peak Gaussian fitting algorithm is expressed as:

[0023]

[0024] Wherein, j represents the subscript number of each peak value in the current cross-section trajectory data, j = 1, 2, 3, ...; y jmax represents the peak value of the jth peak in the current section, x ji represents the i-th coordinate point of the j-th peak in the current section, x jmax represents the peak position of the jth peak in the current section, S j Represents the half-width information of each peak.

[0025] Furthermore, before the step of acquiring the position of the section lane line corresponding to the section trajectory distribution curve in the section lane line distribution sample library according to the similarity algorithm, the following steps are also included:

[0026] Determine the number of lanes;

[0027] Determine the lane starting position set;

[0028] Construct a lane width collection;

[0029] The lane number, the lane starting position set, and the lane width set are combined to generate the cross-section lane line distribution sample library.

[0030] In order to solve the above technical problems, the embodiment of the present application further provides a road lane line generation device, which adopts the following technical solution:

[0031] A data acquisition module, used to receive the original vehicle trajectory data collected by the radar collection device;

[0032] A denoising module, used for performing a denoising operation on the original vehicle trajectory data to obtain denoised vehicle trajectory data;

[0033] A segment division module, used for performing a segment division operation on the denoised vehicle trajectory data to obtain N segment division trajectory data, wherein N is an integer greater than or equal to 1;

[0034] A Gaussian function fitting module is used to perform Gaussian function fitting operations on the N segment-divided trajectory data respectively to obtain N fitting result parameters;

[0035] A linear function fitting module is used to calculate the variance values ​​of the N fitting result parameters and perform a linear function fitting operation on the variance values ​​to obtain a measurement error variance;

[0036] A Gaussian distribution fitting module, used for extracting the peak value of each lane in the measurement error variance according to a multi-peak Gaussian fitting algorithm, and performing Gaussian distribution fitting on each of the peak values ​​to obtain a cross-sectional trajectory distribution curve, wherein the trajectory distribution curve is composed of Gaussian functions of each peak of the multi-peak Gaussian function;

[0037] A position acquisition module, used for acquiring the position of the section lane line corresponding to the section trajectory distribution curve in the section lane line distribution sample library according to a similarity algorithm;

[0038] The lane line generation module is used to generate the target road lane line according to the section lane line position.

[0039] Furthermore, the denoising module includes:

[0040] The correction submodule is used to correct the original vehicle trajectory data according to the data correction algorithm to obtain the denoised vehicle trajectory data, wherein the data correction algorithm is expressed as:

[0041]

[0042]

[0043] in, and They represent the lateral distance and radial distance of the target detected by the radar, respectively; x and y represent the corrected lateral distance and radial distance, respectively; and β represents the angle to be rotated.

[0044] Furthermore, the device also includes:

[0045] The smoothing processing module is used to smooth the measurement error variance according to the Kalman filter algorithm.

[0046] In order to solve the above technical problems, the embodiment of the present application further provides a computer device, which adopts the following technical solution:

[0047] It includes a memory and a processor, wherein the memory stores computer-readable instructions, and when the processor executes the computer-readable instructions, the steps of the road lane line generation method as described above are implemented.

[0048] In order to solve the above technical problems, the embodiment of the present application further provides a computer-readable storage medium, which adopts the following technical solution:

[0049] The computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the steps of the road lane line generation method as described above are implemented.

[0050] The present application provides a method for generating road lane lines, comprising: receiving original vehicle trajectory data collected by a radar acquisition device; performing a denoising operation on the original vehicle trajectory data to obtain denoised vehicle trajectory data; performing a segmentation operation on the denoised vehicle trajectory data to obtain N segmented trajectory data, wherein N is an integer greater than or equal to 1; performing a Gaussian function fitting operation on the N segmented trajectory data respectively to obtain N fitting result parameters; calculating the variance values ​​of the N fitting result parameters, and performing a linear function fitting operation on the variance values ​​to obtain a measurement error variance; extracting the peak values ​​of each lane in the measurement error variance according to a multi-peak Gaussian fitting algorithm, and performing a Gaussian distribution fitting on each of the peak values ​​to obtain a cross-sectional trajectory distribution curve, wherein the trajectory distribution curve is composed of Gaussian functions of each peak of the multi-peak Gaussian function; obtaining a cross-sectional lane line position corresponding to the cross-sectional trajectory distribution curve in a cross-sectional lane line distribution sample library according to a similarity algorithm; and generating a target road lane line according to the cross-sectional lane line position. Compared with the prior art, the present application uses millimeter-wave radar to accurately collect real-time vehicle trajectory data at each road section, divides the road section into multiple sections along the longitudinal section, and performs vehicle operation trajectory distribution statistical analysis on the trajectory data of each single section, and fits the probability density function of the vehicle trajectory of each lane. According to the objective road traffic regulations and the subjective driving habits of the driver, the probability of a vehicle traveling in the lane area is much greater than the probability of a vehicle crossing the lane line, and thus the center point of the lane is obtained based on the probability density function; according to the coordinates of the outermost trajectory point and the lane width according to the urban road intersection planning specifications , the virtual lane line position intervals and the national standard lane widths are arranged and combined, and then a channelization standard library is constructed, thereby obtaining the road lane line positions under different outermost lane line positions and different lane widths. Finally, a similarity algorithm is used to find the channelization scheme with the highest similarity between the center lines of each lane and the center lines of each lane in this section in the channelization standard library. The lane line position of this scheme is the lane line position of the section; finally, the lane lines of each segmented section are continuously fitted to obtain the actual lane line position of the section, which effectively solves the disadvantages of the traditional lane channelization drawing method that is time-consuming and labor-intensive, and has huge labor costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the scheme in the present application, a brief introduction is given below to the drawings required for use in the description of the embodiments of the present application. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0052] Figure 1 is an exemplary system architecture diagram to which the present application may be applied;

[0053] Figure 2 is a flow chart of the implementation of the road lane line generation method provided in Example 1 of the present application;

[0054] Figure 3 This is a schematic diagram of the vertical installation of the millimeter wave radar provided in Example 1 of the present application;

[0055] Figure 4 This is a schematic diagram of the horizontal installation of the millimeter wave radar provided in Example 1 of the present application;

[0056] Figure 5 This is a schematic diagram of the sliding window frame selection section data provided in the first embodiment of the present application;

[0057] Figure 6 This is a schematic diagram of generating continuous road lane lines provided in Example 1 of the present application;

[0058] Figure 7 is a schematic diagram of the structure of a road lane line generating device provided in Embodiment 2 of the present application;

[0059] Figure 8 It is a structural diagram of an embodiment of a computer device according to the present application. DETAILED DESCRIPTION

[0060] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of the present application; the terms used in the specification of the application herein are only for the purpose of describing specific embodiments and are not intended to limit the present application; the terms "including" and "having" and any variations thereof in the specification and claims of the present application and the above-mentioned drawings are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of the present application or the above-mentioned drawings are used to distinguish different objects, not to describe a specific order.

[0061] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0062] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.

[0063] like Figure 1As shown, the system architecture 100 may include terminal devices 101, 102, 103, a network 104 and a server 105. The network 104 is used to provide a medium for communication links between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0064] Users can use terminal devices 101, 102, 103 to interact with server 105 through network 104 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 101, 102, 103, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0065] Terminal devices 101, 102, 103 can be various electronic devices with display screens and supporting web browsing, including but not limited to smart phones, tablet computers, e-book readers, MP3 players (Moving Picture Experts Group Audio Layer III), MP4 (Moving Picture Experts Group Audio Layer IV) players, laptop computers, desktop computers, etc.

[0066] The server 105 may be a server that provides various services, such as a background server that provides support for web pages displayed on the terminal devices 101 , 102 , and 103 .

[0067] It should be noted that the road lane line generation method provided in the embodiment of the present application is generally executed by a server / terminal device, and accordingly, the road lane line generation device is generally set in the server / terminal device.

[0068] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0069] Continue to refer Figure 2 , shows a flow chart of an embodiment of a road lane line generation method according to the present application. The road lane line generation method comprises: step S201, step S202, step S203, step S204, step S205, step S206, step S207 and step S208.

[0070] In step S201, original vehicle trajectory data collected by a radar collection device is received.

[0071] In the embodiment of the present application, the detection area of ​​the system is mainly accident-prone areas such as highway ramp merging areas, diverging areas, tunnels, bridges, etc. Pole poles are set up on each road section, and two-way or even multi-directional millimeter wave radars are installed to achieve full coverage of potential dangerous sections. The installation method is as follows Figure 3 , Figure 4 shown.

[0072] In step S202, a denoising operation is performed on the original vehicle trajectory data to obtain denoised vehicle trajectory data.

[0073] In an embodiment of the present application, after the original data is parsed, the vehicle data is cleaned and corrected based on a multi-strategy comprehensive preprocessing method such as target data screening, filtering and correction to eliminate invalid information such as background noise and false alarm targets in the original traffic data and improve the accuracy of target recognition.

[0074] In the embodiment of the present application, the denoising operation may be a target data screening operation or a target data filtering operation, wherein:

[0075] (1) Target data screening operation: According to the radar data interface protocol manual, the data returned by the radar contains various types of information, such as system status information, software configuration status information, hardware status information, and traffic detection information. Therefore, it is necessary to extract the vehicle target information from the raw data returned by the radar;

[0076] (2) Target data filtering operation: False detection may occur during radar operation. These false detections will reduce the accuracy of valid targets. Therefore, it is necessary to filter out these worthless invalid targets. Specifically:

[0077] ① When the detection vehicle target ID appears less than 5 times in consecutive adjacent sampling cycles, it can be judged as an invalid target and filtered out;

[0078] ② The sampling period of the millimeter-wave radar is set to be short, so the difference between two consecutive target values ​​will not be too large. When the difference between the regional positions of the same target collected twice in a row is too large or the value of the lateral radial velocity exceeds the normal speed threshold, the target data detected this time is judged to be invalid.

[0079] In the embodiment of the present application, the above methods are used to perform comprehensive preprocessing on the vehicle target based on product interface protocol information, motion state information, data change information, regional location and vehicle speed. Data cleaning and filtering are completed, environmental interference to the radar is reduced, false alarms and noise are reduced, the accuracy of vehicle trajectory recognition is improved, and stable support is provided for subsequent data processing.

[0080] In step S203, the denoised vehicle trajectory data is segmented to obtain N segmented trajectory data, where N is an integer greater than or equal to 1.

[0081] In the embodiment of the present application, the segment division operation refers to dividing the vehicle position data collected by the radar into multiple segments equally according to the radial distance, and each segment can be 5 to 20 meters.

[0082] In step S204, Gaussian function fitting operations are performed on the N segment-divided trajectory data to obtain N fitting result parameters.

[0083] In the embodiment of the present application, according to objective road traffic regulations and subjective driving habits of drivers, the probability of a vehicle driving in the lane area is much greater than the probability of a vehicle crossing the lane line, so the lateral distance corresponding to the extreme point should be located at the center of the lane. Therefore, the pre-processed millimeter-wave radar data of each section in the segment is counted. Figure 5 The sliding window method shown is used to select cross-section statistical data. When the extreme point of the data in the window is located in the middle of the window horizontal coordinate, a Gaussian function is fitted to the data in the window. The Gaussian function formula is shown below:

[0084]

[0085] In step S205, the variance values ​​of the N fitting result parameters are calculated, and a linear function fitting operation is performed on the variance values ​​to obtain the measurement error variance.

[0086] In the embodiment of the present application, the fitting result parameters are obtained based on the Gaussian function formula of the statistical number points of each section, and then the value of the function variance is obtained through the fitting parameters, and the radar measurement error is expressed in the form of Gaussian function variance. The relationship between the radar measurement error variance value and the measurement distance basically conforms to the linear relationship. The linear function is fitted to the statistical variance data points, and the measurement error variance obtained by fitting and the radial distance of the target in each section satisfy the linear relationship:

[0087] σ 2 =P×y i (k)-b

[0088] In step S206, the peak values ​​of each lane in the measurement error variance are extracted according to the multi-peak Gaussian fitting algorithm, and Gaussian distribution fitting is performed on each peak to obtain a cross-sectional trajectory distribution curve, wherein the trajectory distribution curve is composed of Gaussian functions of each peak of the multi-peak Gaussian function.

[0089] In an embodiment of the present application, a road transverse section includes multiple lanes, and the vehicle trajectory distribution curve of each lane basically conforms to the Gaussian distribution. Therefore, in order to obtain the trajectory distribution law of the entire section, a multi-peak Gaussian fitting algorithm can be used to extract the peak value of each lane in the data, and a Gaussian distribution fitting can be performed on each peak.

[0090] The multi-peak Gaussian function of is as follows:

[0091]

[0092] Wherein, j represents the subscript number of each peak value in the current cross-section trajectory data, j = 1, 2, 3, ...; yj max Represents the peak value of the jth peak in the current section, xj i represents the i-th coordinate point of the j-th peak in the current section, x jmax represents the peak position of the jth peak in the current section, S j Represents the half-width information of each peak.

[0093] To solve the multi-peak Gaussian function, each peak can be calculated according to the single-peak Gaussian function fitting process, and then the characteristic parameter y to be estimated of the Gaussian function is calculated. max , x max , the value of S.

[0094] Thus, the Gaussian functions of each peak of the multi-peak Gaussian function are obtained, and finally combined into the trajectory distribution curve of the section, x max The value of is the center line position of each lane. max The average value of is the lane line between the two lanes.

[0095] In the embodiment of the present application, the trajectory data will present different measurement errors at different detection radial distances. By analyzing the vehicle trajectory data after segmentation processing, the relationship between the radar measurement error and the measurement distance is linearized, and then filtering processing based on the linear relationship between error and distance is completed to obtain a trajectory distribution function that is easy to extract lane lines.

[0096] In step S207, the cross-section lane line position corresponding to the cross-section trajectory distribution curve is obtained in the cross-section lane line distribution sample library according to the similarity algorithm.

[0097] In an embodiment of the present application, a group of data with the highest correlation with the centerline position and lane line position of each lane obtained from the cross-section trajectory distribution curve is found in the cross-section lane line distribution sample library through a similarity algorithm. Each center line of the group is the center line of each lane of the section, and the lane line position corresponding to the group is the lane line position of each lane of the section.

[0098] In the embodiment of the present application, the position of the lane line of the section may be determined by:

[0099] ① Obtain the centerline position and lane line position of each lane in the current section: In step 1, the vehicle trajectory data is processed according to the road section segmentation, and the x value of each peak of the multi-peak Gaussian function is obtained. max The value of is the center line position of each lane. max The average value of is the lane line between the two lanes.

[0100] ②Similarity algorithm selects the best sample: First determine the data format, and arrange the lane geometry information of the cross section and the lane geometry information of the sample library in the order of the horizontal coordinate of the left lane line, the horizontal coordinate of the middle lane line, and the horizontal coordinate of the right lane line. For example:

[0101] (x left1 ,x mid1 ,x right1 ,x left2 ,x mid2 ,x right2 ,…,x leftn ,x midn ,x rightn )

[0102] Then, the geometric information dataset of the lanes in the sample library is traversed and the geometric information dataset of the lanes in the current section is substituted into the Euclidean distance similarity calculation formula, which is as follows:

[0103]

[0104] The smaller the distance, the greater the similarity. Therefore, the sample that is most similar to the lane line distribution position obtained based on the trajectory distribution law of the current section can be found in the lane line distribution sample library. The lane line position of this sample is the lane line position of the current section.

[0105] In step S208, target road lane lines are generated according to the cross-section lane line positions.

[0106] In the embodiment of the present application, the lane line position distribution (x lane1 , x lane2 , …, x lanei , …x lanen ), arrange them in order according to the longitudinal position of each distribution set, and then perform curve fitting in the longitudinal direction of the road to generate continuous road lane lines, such as Figure 6 shown.

[0107] In an embodiment of the present application, a road lane line generation method is provided, including: receiving original vehicle trajectory data collected by a radar acquisition device; performing a denoising operation on the original vehicle trajectory data to obtain denoised vehicle trajectory data; performing a segmentation operation on the denoised vehicle trajectory data to obtain N segmented trajectory data, wherein N is an integer greater than or equal to 1; performing a Gaussian function fitting operation on the N segmented trajectory data respectively to obtain N fitting result parameters; calculating the variance values ​​of the N fitting result parameters, and performing a linear function fitting operation on the variance values ​​to obtain a measurement error variance; extracting the peak value of each lane in the measurement error variance according to a multi-peak Gaussian fitting algorithm, and performing a Gaussian distribution fitting on each peak to obtain a cross-sectional trajectory distribution curve, wherein the trajectory distribution curve is composed of Gaussian functions of each peak of the multi-peak Gaussian function; obtaining a cross-sectional lane line position corresponding to the cross-sectional trajectory distribution curve in a cross-sectional lane line distribution sample library according to a similarity algorithm; and generating a target road lane line according to the cross-sectional lane line position. Compared with the prior art, the present application uses millimeter-wave radar to accurately collect real-time vehicle trajectory data at each road section, divides the road section into multiple sections along the longitudinal section, and performs vehicle operation trajectory distribution statistical analysis on the trajectory data of each single section, and fits the probability density function of the vehicle trajectory of each lane. According to the objective road traffic regulations and the subjective driving habits of the driver, the probability of a vehicle traveling in the lane area is much greater than the probability of a vehicle crossing the lane line, and thus the center point of the lane is obtained based on the probability density function; according to the coordinates of the outermost trajectory point and the lane width according to the urban road intersection planning specifications , the virtual lane line position intervals and the national standard lane widths are arranged and combined, and then a channelization standard library is constructed, thereby obtaining the road lane line positions under different outermost lane line positions and different lane widths. Finally, a similarity algorithm is used to find the channelization scheme with the highest similarity between the center lines of each lane and the center lines of each lane in this section in the channelization standard library. The lane line position of this scheme is the lane line position of the section; finally, the lane lines of each segmented section are continuously fitted to obtain the actual lane line position of the section, which effectively solves the disadvantages of the traditional lane channelization drawing method that is time-consuming and labor-intensive, and has huge labor costs.

[0108] In some optional implementations of this embodiment, step S202 specifically includes the following steps:

[0109] The original vehicle trajectory data is corrected according to the data correction algorithm to obtain the denoised vehicle trajectory data, where the data correction algorithm is expressed as:

[0110]

[0111]

[0112] in, and They represent the lateral distance and radial distance of the target detected by the radar, respectively; x and y represent the corrected lateral distance and radial distance, respectively; and β represents the angle to be rotated.

[0113] In the embodiment of the present application, the denoising operation can also be a target data correction operation, where, due to equipment installation problems, there is an angle error between the normal line of the radar wave emission surface and the lane direction, causing the vehicle trajectory to deviate from the positive direction. Therefore, it is necessary to correct the original vehicle trajectory data, that is, to correct the two-dimensional coordinate position of the target position point.

[0114] In an embodiment of the present application, a comprehensive pre-processing based on position correction is performed on the vehicle target through a target data correction operation to complete data cleaning and filtering, further reduce environmental interference to the radar, reduce false alarms and noise, improve the accuracy of vehicle trajectory recognition and provide stable support for subsequent data processing.

[0115] In some optional implementations of this embodiment, after step S205 and before step S206, the following steps are further included:

[0116] The measurement error variance is smoothed according to the Kalman filter algorithm.

[0117] In an embodiment of the present application, according to the linear relationship between the measurement error variance and the radial distance of the target in each section, the millimeter-wave radar detection error change function can obtain its covariance matrix, and then substitute it into the filter gain matrix of the Kalman filter algorithm to complete the adaptive improvement of the Kalman filter algorithm based on the millimeter-wave radar measurement error change function, thereby reducing the impact of data fluctuations on the subsequent fitting of the trajectory distribution law.

[0118] In summary, the present application realizes vehicle trajectory data perception based on millimeter wave radar, which can accurately obtain the real-time position information of vehicles in the traffic flow of the road section, and the recognition of lane lines based on real trajectory data can improve the accuracy of lane line position data; the present application proposes a data cleaning and filtering noise reduction process for the vehicle target motion trajectory based on a multi-strategy comprehensive data preprocessing method and an improved Kalman filter algorithm, quantifies the radar measurement error, and performs filtering and smoothing processing on the vehicle trajectory data. Based on the radar acquisition radial distance segmentation, the radar measurement error is quantified and fitted to obtain the measurement error change function, and the Kalman filter algorithm is adaptively improved. Finally, the trajectory data is processed by the improved Kalman filter algorithm to obtain a more accurate, real and stable trajectory result; the present application designs a cross-section vehicle trajectory distribution model based on a multi-peak Gaussian distribution based on the distribution law of vehicle trajectory data on the road cross section, and realizes the recognition of cross-section lane lines based on this model, and due to the continuous accumulation of cross-section trajectory data, the lane line position obtained by the cross-section trajectory distribution law model will be more and more consistent with the real position, greatly improving the accuracy of cross-section lane line recognition.

[0119] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0120] The basic technologies of artificial intelligence generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. Artificial intelligence software technology mainly includes computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning. A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through computer-readable instructions, and the computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it may include the processes of the embodiments of the above-mentioned methods. Among them, the aforementioned storage medium may be a non-volatile storage medium such as a disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0121] It should be understood that, although the steps in the flowchart of the accompanying drawings are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0122] Embodiment 2

[0123] Further references Figure 7 , as a response to the above Figure 2 The present application provides an embodiment of a road lane line generating device, and the device embodiment is similar to Figure 2 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0124] like Figure 7 As shown, the road lane line generating device 200 of this embodiment includes: a data acquisition module 210, a denoising module 220, a segment division module 230, a Gaussian function fitting module 240, a linear function fitting module 250, a Gaussian distribution fitting module 260, a position acquisition module 270 and a lane line generating module 280, wherein:

[0125] The data acquisition module 210 is used to receive the original vehicle trajectory data collected by the radar collection device;

[0126] A denoising module 220, configured to perform a denoising operation on the original vehicle trajectory data to obtain denoised vehicle trajectory data;

[0127] A segment division module 230 is used to perform a segment division operation on the denoised vehicle trajectory data to obtain N segment division trajectory data, where N is an integer greater than or equal to 1;

[0128] A Gaussian function fitting module 240 is used to perform Gaussian function fitting operations on the N segment-divided trajectory data to obtain N fitting result parameters;

[0129] The linear function fitting module 250 is used to calculate the variance values ​​of the N fitting result parameters and perform a linear function fitting operation on the variance values ​​to obtain the measurement error variance;

[0130] A Gaussian distribution fitting module 260 is used to extract the peak value of each lane in the measurement error variance according to a multi-peak Gaussian fitting algorithm, and perform Gaussian distribution fitting on each peak to obtain a cross-sectional trajectory distribution curve, wherein the trajectory distribution curve is composed of Gaussian functions of each peak of the multi-peak Gaussian function;

[0131] A position acquisition module 270, for acquiring a section lane line position corresponding to a section trajectory distribution curve in a section lane line distribution sample library according to a similarity algorithm;

[0132] The lane line generation module 280 is used to generate the target road lane line according to the section lane line position.

[0133] In this embodiment, a road lane line generation device 200 is provided, including: a data acquisition module 210, a denoising module 220, a segment division module 230, a Gaussian function fitting module 240, a linear function fitting module 250, a Gaussian distribution fitting module 260, a position acquisition module 270 and a lane line generation module 280. Compared with the prior art, the present application uses millimeter wave radar to accurately collect real-time vehicle trajectory data at the positions of each road section, divides the road section into multiple sections along the longitudinal section, and performs vehicle operation trajectory distribution statistical analysis on the trajectory data of each single section, and fits the probability density function of the vehicle trajectory of each lane. According to the objective road traffic regulations and the subjective driving habits of the driver, the probability of a vehicle traveling in the lane area is much greater than the probability of a vehicle crossing the lane line, so the center point of the lane is obtained based on the probability density function; according to the coordinates of the outermost trajectory point and the lane width according to the urban road intersection planning specifications. , the virtual lane line position intervals and the national standard lane widths are arranged and combined, and then a channelization standard library is constructed, thereby obtaining the road lane line positions under different outermost lane line positions and different lane widths. Finally, a similarity algorithm is used to find the channelization scheme with the highest similarity between the center lines of each lane and the center lines of each lane in this section in the channelization standard library. The lane line position of this scheme is the lane line position of the section; finally, the lane lines of each segmented section are continuously fitted to obtain the actual lane line position of the section, which effectively solves the disadvantages of the traditional lane channelization drawing method that is time-consuming and labor-intensive, and has huge labor costs.

[0134] In some optional implementations of this embodiment, the denoising module 220 includes:

[0135] The correction submodule is used to correct the original vehicle trajectory data according to the data correction algorithm to obtain denoised vehicle trajectory data, where the data correction algorithm is expressed as:

[0136]

[0137]

[0138] in, and They represent the lateral distance and radial distance of the target detected by the radar, respectively; x and y represent the corrected lateral distance and radial distance, respectively; and β represents the angle to be rotated.

[0139] In some optional implementations of this embodiment, the road lane line generating device 200 further includes: a smoothing processing module, wherein:

[0140] The smoothing processing module is used to smooth the measurement error variance according to the Kalman filter algorithm.

[0141] To solve the above technical problems, the present application also provides a computer device. Figure 8 , Figure 8 This is a basic structural block diagram of the computer device in this embodiment.

[0142] The computer device 300 includes a memory 310, a processor 320, and a network interface 330 that are interconnected and communicated through a system bus. It should be noted that the figure only shows a computer device 300 having components 310-330, but it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (Application Specific Integrated Circuit, ASIC), programmable gate arrays (Field-Programmable Gate Array, FPGA), digital processors (Digital Signal Processor, DSP), embedded devices, etc.

[0143] The computer device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The computer device may interact with a user through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device.

[0144] The memory 310 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 310 can be an internal storage unit of the computer device 300, such as a hard disk or memory of the computer device 300. In other embodiments, the memory 310 can also be an external storage device of the computer device 300, such as a plug-in hard disk equipped on the computer device 300, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card, etc. Of course, the memory 310 can also include both the internal storage unit of the computer device 300 and its external storage device. In this embodiment, the memory 310 is generally used to store the operating system and various application software installed on the computer device 300, such as computer-readable instructions of the road lane line generation method, etc. In addition, the memory 310 can also be used to temporarily store various data that have been output or will be output.

[0145] The processor 320 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips in some embodiments. The processor 320 is generally used to control the overall operation of the computer device 300. In this embodiment, the processor 320 is used to run computer-readable instructions or process data stored in the memory 310, such as computer-readable instructions for running the road lane line generation method.

[0146] The network interface 330 may include a wireless network interface or a wired network interface. The network interface 330 is generally used to establish a communication connection between the computer device 300 and other electronic devices.

[0147] The computer equipment provided by the present application uses millimeter-wave radar to accurately collect real-time vehicle trajectory data at each section of the road, divides the section into multiple sections along the longitudinal section, and performs vehicle operation trajectory distribution statistical analysis on the trajectory data of each single section, and fits the probability density function of the vehicle trajectory of each lane. According to the objective road traffic regulations and the subjective driving habits of the driver, the probability of a vehicle traveling in the lane area is much greater than the probability of a vehicle crossing the lane line, and thus the center point of the lane is obtained based on the probability density function; according to the coordinates of the outermost trajectory point and the lane width according to the urban road intersection planning regulations The virtual lane line position intervals and the national standard lane widths are arranged and combined to construct a channelization standard library, thereby obtaining the road lane line positions under different outermost lane line positions and different lane widths. Finally, a similarity algorithm is used to find the channelization scheme with the highest similarity between the center lines of each lane and the center lines of each lane in this section in the channelization standard library. The lane line position of this scheme is the lane line position of the section. Finally, the lane lines of each segmented section are fitted with continuous lane lines to obtain the actual lane line position of the section, which effectively solves the disadvantages of the traditional lane channelization drawing method that is time-consuming, labor-intensive, and has huge labor costs.

[0148] The present application also provides another embodiment, namely, providing a computer-readable storage medium, which stores computer-readable instructions, and the computer-readable instructions can be executed by at least one processor to enable the at least one processor to perform the steps of the road lane line generation method as described above.

[0149] The computer-readable storage medium provided by the present application uses millimeter-wave radar to accurately collect real-time vehicle trajectory data at each road section, divides the road section into multiple sections along the longitudinal section, and performs vehicle operation trajectory distribution statistical analysis on the trajectory data of each single section, and fits the probability density function of the vehicle trajectory of each lane. According to the objective road traffic regulations and the subjective driving habits of the driver, the probability of a vehicle traveling in the lane area is much greater than the probability of a vehicle crossing the lane line, thereby obtaining the center point of the lane based on the probability density function; according to the coordinates of the outermost trajectory point and the vehicle according to the urban road intersection planning regulations, the center point of the lane is obtained. The virtual lane line position interval and the national standard lane width are arranged and combined to construct a channelization standard library, thereby obtaining the road lane line position under different outermost lane line positions and different lane widths. Finally, the similarity algorithm is used to find the channelization scheme with the highest similarity between the center line of each lane and the center line of each lane of this section in the channelization standard library. The lane line position of this scheme is the lane line position of the section. Finally, the lane lines of each segmented section are continuously fitted to obtain the actual lane line position of the section, which effectively solves the disadvantages of the traditional lane channelization drawing method that is time-consuming and labor-intensive, and has huge labor costs.

[0150] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0151] Obviously, the embodiments described above are only some embodiments of the present application, rather than all embodiments. The preferred embodiments of the present application are given in the accompanying drawings, but they do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present application more thorough and comprehensive. Although the present application is described in detail with reference to the aforementioned embodiments, for those skilled in the art, it is still possible to modify the technical solutions recorded in the aforementioned specific implementation methods, or to perform equivalent replacement of some of the technical features therein. Any equivalent structure made using the contents of the specification and drawings of this application, directly or indirectly used in other related technical fields, is similarly within the scope of patent protection of this application.

Claims

1. A method for generating road lane lines, characterized in that: The steps include: Receive the original vehicle trajectory data collected by the radar collection device; Performing a denoising operation on the original vehicle trajectory data to obtain denoised vehicle trajectory data; Performing a segment division operation on the denoised vehicle trajectory data to obtain N segment division trajectory data, where N is an integer greater than or equal to 1; Performing Gaussian function fitting operations on the N segment division trajectory data respectively to obtain N fitting result parameters; Calculating the variance values ​​of the N fitting result parameters, and performing a linear function fitting operation on the variance values ​​to obtain a measurement error variance; Extracting the peak value of each lane in the measurement error variance according to a multi-peak Gaussian fitting algorithm, and performing Gaussian distribution fitting on each peak to obtain a cross-sectional trajectory distribution curve, wherein the trajectory distribution curve is composed of Gaussian functions of each peak of the multi-peak Gaussian function; Acquire the cross-section lane line position corresponding to the cross-section trajectory distribution curve in the cross-section lane line distribution sample library according to the similarity algorithm; Generate a target road lane line according to the cross-section lane line position.

2. The road lane line generation method according to claim 1, characterized in that: The step of performing a denoising operation on the original vehicle trajectory data to obtain denoised vehicle trajectory data specifically comprises the following steps: The original vehicle trajectory data is corrected according to a data correction algorithm to obtain the denoised vehicle trajectory data, wherein the data correction algorithm is expressed as: in, and They represent the lateral distance and radial distance of the target detected by the radar, respectively; x and y represent the corrected lateral distance and radial distance, respectively; and β represents the angle to be rotated.

3. The road lane line generation method according to claim 1, characterized in that: After the step of calculating the variance values ​​of the N fitting result parameters and performing a linear function fitting operation on the variance values ​​to obtain the measurement error variance, and before the step of extracting the peak values ​​of each lane in the measurement error variance according to a multi-peak Gaussian fitting algorithm and performing Gaussian distribution fitting on each peak value to obtain a cross-sectional trajectory distribution curve, the following steps are also included: The measurement error variance is smoothed according to a Kalman filter algorithm.

4. The road lane line generation method according to claim 1, characterized in that: The multi-peak Gaussian fitting algorithm is expressed as: Wherein, j represents the subscript number of each peak value in the current cross-section trajectory data, j = 1, 2, 3, ...; y jmax represents the peak value of the jth peak in the current section, x ji represents the i-th coordinate point of the j-th peak in the current section, x jmax represents the peak position of the jth peak in the current section, S j Represents the half-width information of each peak.

5. The road lane line generation method according to claim 1, characterized in that: Before the step of acquiring the position of the cross-section lane line corresponding to the cross-section trajectory distribution curve in the cross-section lane line distribution sample library according to the similarity algorithm, the following steps are also included: Determine the number of lanes; Determine the lane starting position set; Construct a lane width collection; The lane number, the lane starting position set, and the lane width set are combined to generate the cross-section lane line distribution sample library.

6. A road lane line generating device, characterized in that: include: A data acquisition module, used to receive the original vehicle trajectory data collected by the radar collection device; A denoising module, used for performing a denoising operation on the original vehicle trajectory data to obtain denoised vehicle trajectory data; A segment division module, used for performing a segment division operation on the denoised vehicle trajectory data to obtain N segment division trajectory data, wherein N is an integer greater than or equal to 1; A Gaussian function fitting module is used to perform Gaussian function fitting operations on the N segment-divided trajectory data respectively to obtain N fitting result parameters; A linear function fitting module is used to calculate the variance values ​​of the N fitting result parameters and perform a linear function fitting operation on the variance values ​​to obtain a measurement error variance; A Gaussian distribution fitting module, used for extracting the peak value of each lane in the measurement error variance according to a multi-peak Gaussian fitting algorithm, and performing Gaussian distribution fitting on each of the peak values ​​to obtain a cross-sectional trajectory distribution curve, wherein the trajectory distribution curve is composed of Gaussian functions of each peak of the multi-peak Gaussian function; A position acquisition module, used for acquiring the position of the section lane line corresponding to the section trajectory distribution curve in the section lane line distribution sample library according to a similarity algorithm; The lane line generation module is used to generate the target road lane line according to the section lane line position.

7. The road lane line generating device according to claim 6, characterized in that: The denoising module comprises: The correction submodule is used to correct the original vehicle trajectory data according to the data correction algorithm to obtain the denoised vehicle trajectory data, wherein the data correction algorithm is expressed as: in, and They represent the lateral distance and radial distance of the target detected by the radar, respectively; x and y represent the corrected lateral distance and radial distance, respectively; and β represents the angle to be rotated.

8. The road lane line generating device according to claim 6, characterized in that: The device also includes: The smoothing processing module is used to smooth the measurement error variance according to the Kalman filter algorithm.

9. A computer device, comprising a memory and a processor, wherein the memory stores computer-readable instructions, and the processor implements the steps of the road lane line generation method according to any one of claims 1 to 5 when executing the computer-readable instructions.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by a processor, the steps of the road lane line generation method according to any one of claims 1 to 5 are implemented.

Images