Search

CN-121982908-A - Road congestion degree assessment method and device based on laser radar detection

CN121982908ACN 121982908 ACN121982908 ACN 121982908ACN-121982908-A

Abstract

The application discloses a road congestion degree assessment method and device based on laser radar detection, and relates to the field of vehicle control, wherein the method comprises the steps of constructing an ENU coordinate system taking a preset reference point as a coordinate origin as a reference coordinate system, and calculating an external reference relation between a main laser radar point cloud coordinate system and the reference coordinate system; drawing and recording a closed boundary of a selected road section on an XOY projection surface of a reference coordinate system, preprocessing laser radar point clouds, including coordinate conversion, removing point clouds and ground point clouds outside the boundary of the road section, acquiring a laser radar vehicle detection method in deep learning, detecting all vehicles in the selected road section at a certain moment based on the laser radar vehicle detection method in the deep learning, counting the number of the vehicles to obtain vehicle flow data, smoothing vehicle tracks by correlation matching of frames before and after scanning and Kalman filtering, calculating real-time speeds of the vehicles, and counting average and median of all effective vehicle speeds in the road section.

Inventors

  • WU ZAILIN
  • HAN RUI

Assignees

  • 一汽解放汽车有限公司

Dates

Publication Date
20260505
Application Date
20260115

Claims (10)

  1. 1. The road congestion degree assessment method based on laser radar detection is characterized by comprising the following steps of: Step 1, constructing an ENU coordinate system with a preset reference point as a coordinate origin as a reference coordinate system, and calculating an external reference relation between a main laser radar point cloud coordinate system and the reference coordinate system; step 2, drawing and recording a closed boundary of the selected road section on an XOY projection plane of the reference coordinate system; step 3, preprocessing the laser radar point cloud, including coordinate conversion and eliminating point cloud outside the boundary of the road section and ground point cloud; Step 4, acquiring a laser radar vehicle detection method in deep learning; Based on a laser radar vehicle detection method in deep learning, detecting all vehicles in a selected road section at a certain moment and counting the number of the vehicles to obtain vehicle flow data; Step 5, calculating real-time speed of the vehicle by scanning the association of the front frame and the rear frame to match the vehicle and the Kalman filtering smooth vehicle track, and counting the average and median of all effective vehicle speeds in the road section; step 6, obtaining a congestion index through weighted analysis based on the traffic flow data, the vehicle speed average number and the median, and judging the congestion degree; And 7, updating and releasing the road section congestion degree evaluation result according to the preset frequency.
  2. 2. The method for evaluating the congestion degree of a road based on laser radar detection according to claim 1, wherein the step 2 comprises: obtaining road section boundary points through unmanned plane measurement, laser radar measurement or satellite map measurement; Wherein, a closed boundary is constructed by adopting a curve fitting mode.
  3. 3. The method for evaluating the congestion degree of a road based on laser radar detection according to claim 1, wherein the step 3 comprises: preprocessing the laser radar point cloud: converting the auxiliary laser radar point cloud into a main laser radar coordinate system, splicing, and uniformly converting all the obtained point clouds into an ENU coordinate system; and removing point clouds which are outside the boundary of the road section and are more than a preset height from the ground, and dividing and removing the ground point clouds by adopting a random sampling consensus RANSAC algorithm.
  4. 4. The method for evaluating the congestion level of a road based on laser radar detection according to claim 1, wherein the step 4 comprises: the deep learning model is selected from PointRCNN, pointPillars or a SECOND model; the vehicle detection process comprises the steps of point cloud feature extraction, vehicle target prediction, confidence threshold filtering and non-maximum suppression NMS processing, and then counting the number of vehicles in a road section.
  5. 5. The method for evaluating the congestion degree of a road based on laser radar detection according to claim 1, wherein the step 5 comprises: Adopting a Hungary algorithm or an IOU matching method to realize the association matching of the front and rear frames of the vehicle, and distributing a unique ID for each vehicle; Calculating the real-time speed of the vehicle through the displacement difference of the smooth positions of the adjacent frames and the frame interval time, and eliminating abnormal speed exceeding a reasonable range and vehicle speed data with detection confidence coefficient less than 0.7; Wherein IOU is more than or equal to 0.5.
  6. 6. The method for evaluating the congestion degree of a road based on the detection of laser radar according to claim 3, wherein the step 6 comprises: firstly, carrying out standardization treatment on traffic flow, speed average and median, and mapping to a [0,1] interval; The calculation formula of the congestion index of the weighted analysis is CI=alpha× (1-V_avg_std) +beta× (1-V_med_std) +gamma×F_std; wherein alpha+beta +γ=1.0; V_avg_std is a speed average normalized value; V_med_std is a speed median normalized value; f_std is a traffic flow standardized value; Wherein, the weight coefficients alpha, beta and gamma are dynamically adjusted: Wherein, by default α=0.3, β=0.3, γ=0.4; The highway section is adjusted to be alpha=0.4, beta=0.4 and gamma=0.2; City intersection segments are adjusted to α=0.2, β=0.2, γ=0.6; wherein, the congestion degree is classified according to the congestion index CI: CI <0.2 is smooth, CI <0.2 is less than or equal to 0.4 is light congestion, CI <0.4 is less than or equal to 0.7 is medium congestion, and CI is more than or equal to 0.7 is serious congestion.
  7. 7. A road congestion degree evaluation device based on laser radar detection is characterized in that, the laser radar detection-based road congestion degree evaluation method according to any one of claims 1 to 6, comprising: The system comprises a main laser radar, a plurality of auxiliary laser radars and a bracket, which are arranged at the road side, and computer equipment and a storage medium, wherein the computer equipment and the storage medium are used for processing laser radar point cloud information and completing a detection function; The main laser radar and the auxiliary laser radar are selected, assembled and installed at positions which meet the requirement that the radar field of view completely covers a selected road section representing the road congestion degree to be detected; the laser radar is selected from a mechanical laser radar, a semi-solid laser radar or a solid laser radar; and after the auxiliary laser radar is installed, calibrating the auxiliary laser radar with the main laser radar to obtain external parameters, so as to ensure the unified processing of the point cloud data.
  8. 8. The electronic equipment is characterized by comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory are communicated with each other through the communication bus; A computer program is stored in a memory, which when executed by a processor causes the processor to perform the steps of the method for estimating road congestion based on lidar detection according to any of claims 1 to 6.
  9. 9. A computer-readable storage medium, characterized in that a computer program executable by an electronic device is stored, which, when run on the electronic device, causes the electronic device to perform the steps of the method for estimating a congestion level of a road based on lidar detection according to any of claims 1 to 6.
  10. 10. A vehicle platform, comprising: An electronic device for implementing the steps of the road congestion degree evaluation method based on lidar detection according to any one of claims 1 to 6; A processor that runs a program, the data output from the electronic device when the program is run performing the steps of the road congestion degree evaluation method based on lidar detection according to any one of claims 1 to 6; a storage medium storing a program that, when executed, performs the steps of the road congestion degree evaluation method based on lidar detection according to any one of claims 1 to 6 on data output from an electronic device.

Description

Road congestion degree assessment method and device based on laser radar detection Technical Field The application relates to the field of vehicle control, in particular to a road congestion degree evaluation method based on laser radar detection, a road congestion degree evaluation device based on laser radar detection, electronic equipment, a storage medium and a vehicle platform. Background The accurate evaluation of the congestion condition of the road can help drivers to pre-judge the driving route in advance or help traffic management departments to master traffic conditions. At present, most of the methods for automatically evaluating the road congestion degree use a computer vision mode, and the mode is mostly influenced by light conditions, shielding and the like under outdoor conditions, so that the identification capability is limited. Compared with a camera, the laser radar has the advantages of strong light influence resistance, accurate position measurement and the like, and is more suitable for real-time detection of targets under outdoor open conditions. The congestion degree of the road is estimated by combining the traffic flow of a certain specific representative road section in the road with the traffic speed and other information of the vehicle. The method has the advantages of being accurate in detection, high in real-time performance, high in automation degree and the like. In contrast to a method, apparatus, electronic device, and readable storage medium (CN 202411632935.1) for identifying patent road congestion, the patent uses a method of visual image detection to determine the congestion level by the amount of change in the number of vehicles and the like. The method is the same as the method for detecting the number of vehicles in comparison with the method, but the method does not change the comparison after detecting the number of vehicles, and the method can judge the road congestion degree by representing the vehicle density and the vehicle speed of the road section according to the number of vehicles at a certain moment of a fixed road section. The patent determines a plurality of road segments in a road network database, calculates the traffic flow of each road segment, calculates the congestion index of all selected road segments by combining the congestion index of each selected road segment and the weight contributing to the congestion condition of all selected road segments, and evaluates the congestion degree of the road. Similar to the identification of the number of vehicles on the selected road section in this patent, the judgment of other indexes such as vehicle speed, etc. affecting the degree of congestion is not performed, and the specific road section traffic flow detection mode is not described in this patent. Therefore, a strategy for evaluating the road congestion degree based on laser radar detection is needed, and the laser radar is utilized to detect the real-time vehicle data representing the road section to represent the vehicle flow, so that the laser radar is utilized to detect and update the road congestion degree in real time. Disclosure of Invention The invention aims to provide a road congestion degree evaluation method based on laser radar detection, a road congestion degree evaluation device based on laser radar detection, electronic equipment, a storage medium and a vehicle platform, which at least solve one of a plurality of technical problems. For example, when the traditional computer vision channel evaluates congestion, the traditional computer vision channel is easily influenced by light and shielding, and the outdoor scene recognition capability is limited; for example, in some prior art, congestion is estimated only by means of a single index of traffic flow, and key parameters of the vehicle speed are not included, so that the problem of inaccurate estimation results is caused; for example, in some prior art, a specific detection mode of the traffic flow is not specified, and a technical scheme capable of falling to the ground is not known; For example, the prior art is difficult to realize real-time updating, quantization and grading of the congestion state, and can not meet the requirements of dynamic management and control of a driver prejudging route and a traffic management department. The invention provides the following scheme: according to a first aspect of the present invention, there is provided a road congestion degree evaluation method based on lidar detection, comprising: Step 1, constructing an ENU coordinate system with a preset reference point as a coordinate origin as a reference coordinate system, and calculating an external reference relation between a main laser radar point cloud coordinate system and the reference coordinate system; Step 2, drawing and recording a closed boundary of the selected road section on an XOY projection plane of a reference coordinate system; step 3, preprocessing the laser radar point clou