Traffic management auxiliary perception method and device based on radar image information fusion

Abstract

The present invention discloses a traffic management auxiliary perception method and device based on radar image information fusion, which relates to the technical field of intelligent transportation. The traffic management auxiliary perception method based on radar image information fusion first performs fusion processing on monitoring data and radar image information to perceive, monitor, identify, and track moving targets, then judges the traffic density coefficient, then judges the traffic update coefficient, and finally performs proximity monitoring and congestion warning. It can achieve a comprehensive perception and real-time monitoring of the traffic condition, and perform warning and management according to the traffic density coefficient and update situation, which helps to improve traffic safety and efficiency. It solves the problem that the state perception by simply using traffic data cannot solve the influence of various road conditions, traffic congestion, and traffic accidents on the traffic state, thus greatly reducing the effect of traffic state perception and affecting the judgment of the state.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent transportation, and in particular to a traffic management auxiliary perception method and device based on radar image information fusion. Background Art

[0002] With the rapid development of road traffic and the continuous improvement of urban road networks, great progress has been made in traffic infrastructure damage reduction. Along with the improvement of people's living standards, cars have gradually become essential for some families to travel.

[0003] With the increase in the number of motor vehicles, traffic congestion, traffic accidents, and traffic violations occur frequently. Traffic information is the available information for motor vehicles to circulate in the transportation field. Various driving states and driving data of motor vehicles are stored in traffic information. However, in the prior art, when using traffic information for traffic state perception, most of them use existing traffic data for clustering and classify the clustering results according to a pre-constructed analysis model to judge the traffic state. However, simply using traffic data for state perception cannot solve the influence of various road conditions, traffic congestion, and traffic accidents on the traffic state, thus greatly reducing the effect of traffic state perception and affecting the judgment of the state. Summary of the Invention

[0004] Aiming at the deficiencies of the prior art, the present invention provides a traffic management auxiliary perception method and device based on radar image information fusion, which solves the problem that simply using traffic data for state perception cannot solve the influence of various road conditions, traffic congestion, and traffic accidents on the traffic state, thus greatly reducing the effect of traffic state perception and affecting the judgment of the state.

[0005] To achieve the above object, the present invention is implemented by the following technical solutions: A traffic management auxiliary perception method based on radar image information fusion, comprising the following steps: performing fusion processing on monitoring data and radar image information, perceiving and monitoring moving targets in the lane section to be monitored, and identifying and tracking moving targets in the lane section to be monitored; determining the traffic density coefficient of the lane section to be monitored in the current monitoring period based on the identified and tracked moving targets in the lane section to be monitored, and determining whether the traffic density coefficient of the monitored lane section in the current monitoring period is greater than a set first density coefficient threshold. If not, continue to determine whether the traffic density coefficient of the monitored lane section in the next monitoring period is greater than the set first density coefficient threshold. If so, issue a traffic jam warning; obtaining the traffic update coefficient of the lane section to be monitored in the current monitoring period, and determining whether the traffic update coefficient in the current monitoring period is less than the set traffic update threshold coefficient. If so, perform perception monitoring on the adjacent lane section to be monitored. If not, when the traffic density coefficient of the lane section to be monitored in the next monitoring period is greater than the set first density coefficient threshold, determine whether the traffic update coefficient in the next monitoring period is less than the set traffic update threshold coefficient; based on the perception monitoring of the adjacent lane section to be monitored, identifying and tracking moving targets in the adjacent lane section to be monitored, determining the traffic density coefficient of the adjacent lane section to be monitored in the current monitoring period, and determining whether the traffic density coefficient of the adjacent lane section to be monitored in the current monitoring period is greater than the set second density coefficient threshold. If so, issue a first-level severe traffic jam warning. If not, determine whether the traffic update coefficient of the adjacent lane section to be monitored in the current monitoring period is less than the set traffic update threshold coefficient. If so, issue a second-level severe traffic jam warning. If not, when the traffic density coefficient of the lane section to be monitored in the next monitoring period is greater than the set first density coefficient threshold and the traffic density coefficient of the adjacent lane section to be monitored in the next monitoring period is greater than the set second density coefficient threshold, determine whether the traffic update coefficient of the adjacent lane section to be monitored in the next monitoring period is less than the set traffic update threshold coefficient.

[0006] Optionally, the moving targets include moving pedestrians and moving vehicles, and the calculation formula of the traffic density coefficient is: ; where is the traffic density coefficient, is the pedestrian traffic density, and the pedestrian traffic density is determined based on the number of pedestrians in the lane section to be monitored or the number of pedestrians in the adjacent lane section to be monitored, is the set pedestrian traffic density threshold, is the vehicle traffic density, is the set vehicle traffic density threshold, is the pedestrian density weight coefficient, is the vehicle density weight coefficient, and .

[0007] Optionally, the first density coefficient threshold includes a first pedestrian density coefficient threshold, a first vehicle density coefficient threshold, and a first comprehensive density coefficient threshold. The traffic jam warning includes a pedestrian jam warning, a vehicle jam warning, and a comprehensive jam warning. The process of determining whether the traffic density coefficient in the monitored lane section is greater than the set first density coefficient threshold is as follows: When the pedestrian traffic density is greater than the set pedestrian traffic density threshold and the vehicle traffic density is less than the set vehicle traffic density threshold, determine whether the traffic density coefficient is greater than the set first pedestrian density coefficient threshold. If so, issue a pedestrian jam warning; When the pedestrian traffic density is less than the set pedestrian traffic density threshold and the vehicle traffic density is greater than the set vehicle traffic density threshold, determine whether the traffic density coefficient is greater than the set first vehicle density coefficient threshold. If so, issue a vehicle jam warning; When the pedestrian traffic density is greater than the set pedestrian traffic density threshold and the vehicle traffic density is greater than the set vehicle traffic density threshold, determine whether the traffic density coefficient is greater than the set first comprehensive density coefficient threshold. If so, issue a comprehensive jam warning.

[0008] Optionally, the second density coefficient threshold includes a second pedestrian density coefficient threshold, a second vehicle density coefficient threshold, and a second comprehensive density coefficient threshold. The traffic jam warning includes a serious pedestrian jam warning, a serious vehicle jam warning, and a comprehensive serious jam warning. The process of determining whether the traffic density coefficient in the monitored lane section is greater than the set second density coefficient threshold is as follows: When the pedestrian traffic density is greater than the set pedestrian traffic density threshold and the vehicle traffic density is less than the set vehicle traffic density threshold, determine whether the traffic density coefficient is greater than the set second pedestrian density coefficient threshold. If so, issue a serious pedestrian jam warning; When the pedestrian traffic density is less than the set pedestrian traffic density threshold and the vehicle traffic density is greater than the set vehicle traffic density threshold, determine whether the traffic density coefficient is greater than the set second vehicle density coefficient threshold. If so, issue a serious vehicle jam warning; When the pedestrian traffic density is greater than the set pedestrian traffic density threshold and the vehicle traffic density is greater than the set vehicle traffic density threshold, determine whether the traffic density coefficient is greater than the set second comprehensive density coefficient threshold. If so, issue a comprehensive serious jam warning.

[0009] Optionally, the calculation formula for the traffic update coefficient is:

[0010] ;

[0011] In the formula, is the traffic update coefficient, is the pedestrian update rate, is the set pedestrian update reference rate, is the vehicle update rate, is the set vehicle update reference rate, is the weight coefficient of the pedestrian update rate, is the weight coefficient of the vehicle update rate, and , is the natural constant, is the pedestrian traffic density, and the pedestrian traffic density is determined based on the number of pedestrians in the lane section to be monitored or the number of pedestrians in the adjacent monitoring lane section, is the set pedestrian traffic density threshold, is the vehicle traffic density, is the set vehicle traffic density threshold.

[0012] Optionally, the traffic update threshold coefficient includes a pedestrian update threshold coefficient, a vehicle update threshold coefficient, and a comprehensive update threshold coefficient. The process of determining whether the traffic update coefficient is less than the set traffic update threshold coefficient is as follows: When the pedestrian traffic density is greater than the set pedestrian traffic density threshold and the vehicle traffic density is less than the set vehicle traffic density threshold, determine whether the traffic update coefficient is less than the set pedestrian update threshold coefficient. If so, determine whether the traffic density coefficient in the adjacent monitoring lane section is greater than the set second density coefficient threshold. If not, continue to determine whether the traffic update coefficient is less than the set pedestrian update threshold coefficient; When the pedestrian traffic density is less than the set pedestrian traffic density threshold and the vehicle traffic density is greater than the set vehicle traffic density threshold, determine whether the traffic update coefficient is less than the set vehicle update threshold coefficient. If so, determine whether the traffic density coefficient in the adjacent monitoring lane section is greater than the set second density coefficient threshold. If not, continue to determine whether the traffic update coefficient is less than the set vehicle update threshold coefficient; When the pedestrian traffic density is greater than the set pedestrian traffic density threshold and the vehicle traffic density is greater than the set vehicle traffic density threshold, determine whether the traffic update coefficient is less than the set comprehensive update threshold coefficient. If so, determine whether the traffic density coefficient in the adjacent monitoring lane section is greater than the set second density coefficient threshold. If not, continue to determine whether the traffic update coefficient is less than the set comprehensive update threshold coefficient.

[0013] Optionally, the process of perceiving, monitoring, identifying, and tracking a moving target is as follows: Based on the monitoring data, feature extraction is performed on the moving targets entering the lane section to be monitored or the adjacent monitoring lane section, and the pedestrian feature matching dataset and the vehicle feature matching dataset are obtained and stored; Based on the set matching period, feature extraction is performed on the moving targets in the lane section to be monitored or the adjacent monitoring lane section, the pedestrian feature dataset and the vehicle feature dataset are obtained, and the pedestrian feature dataset and the vehicle feature dataset are matched with the pedestrian feature matching dataset and the vehicle feature matching dataset; Based on the matching results, the moving targets are identified and tracked, the moving pedestrians and moving vehicles in the lane section to be monitored are determined, and the moving pedestrians and moving vehicles in the lane section to be monitored are used as the analysis basis for the traffic density coefficient of the monitoring lane section, and the moving pedestrians and moving vehicles in the adjacent monitoring lane section are used as the analysis basis for the traffic density coefficient of the adjacent monitoring lane section.

[0014] Optionally, the pedestrian feature dataset includes a facial feature dataset, a head feature dataset, and a clothing feature dataset. The pedestrian feature matching dataset includes a facial feature matching dataset, a head feature matching dataset, and a clothing feature matching dataset. The pedestrian feature dataset and the pedestrian feature matching dataset are matched to obtain a pedestrian matching coefficient. The pedestrian matching coefficient is used as the basis for identifying and tracking the moving target and also as the analysis basis for the traffic update coefficient; The vehicle feature dataset and the vehicle feature matching dataset are matched to obtain a vehicle matching coefficient. The vehicle matching coefficient is used as the basis for identifying and tracking the moving target and also as the analysis basis for the traffic update coefficient.

[0015] Optionally, the facial feature matching dataset includes the facial occluder area, the occluder curvature, and the occluder bending angle. The head feature dataset includes the distance between the eyebrows, the lip width, and the distance from the upper lip to the eyebrows. The clothing feature dataset includes the upper body width value and the neckline curvature; The facial feature matching dataset includes the facial occluder matching area, the occluder matching curvature, and the occluder matching bending angle. The head feature matching dataset includes the matched distance between the eyebrows, the matched lip width, and the matched distance from the upper lip to the eyebrows. The clothing feature matching dataset includes the upper body matched width value and the neckline matched curvature; If the facial occluder area is less than the set facial occluder area threshold, the pedestrian matching coefficient is obtained based on the head feature dataset, the clothing feature dataset, the head feature matching dataset, and the clothing feature matching dataset. If the facial occluder area is greater than the set facial occluder area threshold, the pedestrian matching coefficient is obtained based on the facial feature dataset, the clothing feature dataset, the facial feature matching dataset, and the clothing feature matching dataset; The vehicle feature dataset includes the windshield area and the vehicle HSV value. The vehicle feature matching dataset includes the windshield matched area and the vehicle HSV matched value.

[0016] A traffic management auxiliary perception method and device based on radar image information fusion, including a moving target perception and monitoring module, a monitoring lane section monitoring and warning module, a traffic update monitoring module for the monitoring lane section, and a monitoring and warning module for the adjacent monitoring lane section, where: The moving target perception and monitoring module is used to fuse and process monitoring data and radar image information, perceive and monitor moving targets in the monitoring lane section, and identify and track moving targets in the monitoring lane section; The monitoring lane section monitoring and warning module is used to determine the traffic density coefficient of the monitoring lane section in the current monitoring period based on the identified and tracked moving targets in the monitoring lane section, and judge whether the traffic density coefficient of the monitoring lane section in the current monitoring period is greater than the set first density coefficient threshold. If not, continue to judge whether the traffic density coefficient of the monitoring lane section in the next monitoring period is greater than the set first density coefficient threshold. If so, issue a traffic jam warning; The traffic update monitoring module for the monitoring lane section is used to obtain the traffic update coefficient of the monitoring lane section in the current monitoring period, and judge whether the traffic update coefficient in the current monitoring period is less than the set traffic update threshold coefficient. If so, perform perception and monitoring of the adjacent monitoring lane section. If not, judge whether the traffic update coefficient in the next monitoring period is less than the set traffic update threshold coefficient when the traffic density coefficient of the monitoring lane section in the next monitoring period is greater than the set first density coefficient threshold; The adjacent monitoring lane section monitoring and warning module is used to identify and track moving targets in the adjacent monitoring lane section based on the perception and monitoring of the adjacent monitoring lane section, determine the traffic density coefficient of the adjacent monitoring lane section in the current monitoring period, and judge whether the traffic density coefficient of the adjacent monitoring lane section in the current monitoring period is greater than the set second density coefficient threshold. If so, issue a first-level serious traffic jam warning. If not, judge whether the traffic update coefficient of the adjacent monitoring lane section in the current monitoring period is less than the set traffic update threshold coefficient. If so, issue a second-level serious traffic jam warning. If not, judge whether the traffic update coefficient of the adjacent monitoring lane section in the next monitoring period is less than the set traffic update threshold coefficient when the traffic density coefficient of the monitoring lane section in the next monitoring period is greater than the set first density coefficient threshold and the traffic density coefficient of the adjacent monitoring lane section in the next monitoring period is greater than the set second density coefficient threshold.

[0017] The technical solution of the present invention has at least the following beneficial effects compared with the prior art:

[0018] In the above solution, by integrating monitoring data and radar image information, and perceiving, monitoring, identifying and tracking moving targets, a comprehensive perception and tracking of the traffic scene is achieved, thus effectively monitoring the traffic condition. Using information such as the density of pedestrians and vehicles, combined with relevant calculation formulas and thresholds, real-time analysis and warning of traffic density are realized, and problems such as traffic congestion can be discovered in time.

[0019] By setting different density coefficient thresholds and update thresholds, and comprehensively considering the matching results of different features, accurate early warning of serious congestion of pedestrians and vehicles is achieved, which helps to take traffic management measures in a timely manner. Multi-dimensional feature matching is adopted, such as the facial features, head features, and clothing features of pedestrians, as well as the windshield area and HSV value of vehicles, etc., which improves the accuracy and reliability of matching. Brief Description of the Drawings

[0020] Figure 1 It is a flowchart of the traffic management auxiliary perception method based on radar image information fusion of the present invention;

[0021] Figure 2 It is a connection diagram of the traffic management auxiliary perception device based on radar image information fusion of the present invention. Detailed Embodiments

[0022] In the embodiment of the present invention, through the traffic management auxiliary perception method based on radar image information fusion, by fusing radar image information and combining monitoring data, comprehensive perception and monitoring of the traffic state are achieved, and early warning and adjustment can be carried out in a timely manner, thereby effectively alleviating traffic congestion, improving road traffic efficiency, and enhancing traffic safety.

[0023] The general idea for the problems in the embodiment of the present invention is as follows:

[0024] First, the monitoring data and radar image information are fused. This may involve steps such as data preprocessing, feature extraction, and data fusion to ensure that the information obtained from two different data sources can be effectively combined. The fused data is used for the perception and monitoring of moving targets. This step involves perception algorithms such as target detection and target recognition, and the position information of moving targets can be extracted from the monitoring data and radar images.

[0025] Based on the identified and tracked moving targets, the traffic density coefficient of the lane section to be monitored is determined. This step can be achieved by statistically analyzing the quantity information of the moving targets to reflect the traffic density of the lane section. The traffic state is judged according to the traffic density coefficient, and corresponding early warnings are made. For example, if the traffic density coefficient exceeds the set threshold, a traffic congestion early warning is issued; if the traffic density coefficient is lower than the set threshold but the traffic update coefficient is low, the perception and monitoring of the adjacent monitored lane section are carried out; if the traffic density coefficient of the adjacent monitored lane section exceeds the second density coefficient threshold, a traffic serious congestion early warning is issued.

[0026] Monitor the traffic update coefficient to timely detect changes in the traffic condition. If the traffic update coefficient is lower than the set threshold, the perception and monitoring of the adjacent monitored lane section are carried out, as well as possible traffic state adjustments.

[0027] As shown Figure 1 in the figure, an embodiment of the present invention provides a traffic management auxiliary perception method based on radar image information fusion, including the following steps: fusing monitoring data with radar image information to perceive and monitor moving targets in the lane section to be monitored, identifying and tracking moving targets in the lane section to be monitored; determining the traffic density coefficient of the lane section to be monitored in the current monitoring period based on the identified and tracked moving targets in the lane section to be monitored, and judging whether the traffic density coefficient of the monitored lane section in the current monitoring period is greater than the set first density coefficient threshold. If not, continue to judge whether the traffic density coefficient of the monitored lane section in the next monitoring period is greater than the set first density coefficient threshold. If so, issue a traffic jam warning; obtaining the traffic update coefficient of the lane section to be monitored in the current monitoring period, and judging whether the traffic update coefficient in the current monitoring period is less than the set traffic update threshold coefficient. If so, perform perception monitoring on the adjacent lane section to be monitored. If not, judge whether the traffic update coefficient in the next monitoring period is less than the set traffic update threshold coefficient when the traffic density coefficient of the lane section to be monitored in the next monitoring period is greater than the set first density coefficient threshold; based on the perception monitoring of the adjacent lane section to be monitored, identifying and tracking moving targets in the adjacent lane section to be monitored, determining the traffic density coefficient of the adjacent lane section to be monitored in the current monitoring period, and judging whether the traffic density coefficient of the adjacent lane section to be monitored in the current monitoring period is greater than the set second density coefficient threshold. If so, issue a first-level serious traffic jam warning. If not, judge whether the traffic update coefficient of the adjacent lane section to be monitored in the current monitoring period is less than the set traffic update threshold coefficient. If so, issue a second-level serious traffic jam warning. If not, judge whether the traffic update coefficient of the adjacent lane section to be monitored in the next monitoring period is less than the set traffic update threshold coefficient when the traffic density coefficient of the lane section to be monitored in the next monitoring period is greater than the set first density coefficient threshold and the traffic density coefficient of the adjacent lane section to be monitored in the next monitoring period is greater than the set second density coefficient threshold.

[0028] The process of fusing monitoring data with radar image information to identify and track moving targets in the lane section to be monitored is as follows: Through cameras installed in the lane section to be monitored, monitor video data is obtained in real time. Through a radar system installed in the same section, information on moving targets detected by the radar is obtained in real time, including data such as the distance, speed, and direction of the targets. The video data and radar data are synchronized in time to ensure that the information of the two data sources is processed on the same time axis. Based on the installation positions and angles of the monitoring cameras and radars, spatial alignment is performed, and the video data and radar data are mapped into the same coordinate system. Image features and radar features are used for matching to associate the targets detected by the radar with the targets in the video, ensuring that the same target has a consistent identifier in the two data sources. A multi-target tracking algorithm is used to process the fused data to track each moving target in real time. Based on the historical positions and motion states of the targets, the future positions of the targets are predicted, and the tracking trajectories are updated. The appearance features (such as shape, color, etc.) and motion features (such as speed, direction, etc.) of the targets are used to ensure the consistency of the identities of the tracked targets. The fused monitoring data and radar information are displayed on the monitoring interface, including the real-time positions, motion trajectories, etc. of each target, and pedestrian feature data and vehicle feature data are obtained through the monitoring data.

[0029] Specifically, the moving targets include moving pedestrians and moving vehicles. In addition to being obtained through relevant monitoring devices or directly represented by the densities of pedestrians and vehicles, the traffic density coefficient can also be calculated through the following formula. The calculation formula for the traffic density coefficient is:

[0030] ;

[0031] In the formula, is the traffic density coefficient, is the pedestrian traffic density, which is determined based on the number of pedestrians in the lane section to be monitored or the number of pedestrians in the adjacent monitored lane section, is the set pedestrian traffic density threshold, is the vehicle traffic density, is the set vehicle traffic density threshold, is the pedestrian density weight coefficient, is the vehicle density weight coefficient, and .

[0032] In this implementation scheme, the impacts of two types of moving targets, pedestrians and vehicles, on traffic density are considered, enabling the traffic density coefficient to more comprehensively reflect the traffic conditions on the road. By dynamically adjusting the pedestrian density weight coefficient and the vehicle density weight coefficient, the calculation of the traffic density coefficient becomes more flexible, capable of being adjusted according to the actual situation and the characteristics of different traffic scenarios, thus adapting to different traffic environments and making the calculation of the traffic density coefficient more universal and adaptable.

[0033] Specifically, the first density coefficient threshold includes the first pedestrian density coefficient threshold, the first vehicle density coefficient threshold, and the first comprehensive density coefficient threshold. The traffic jam warning includes pedestrian jam warning, vehicle jam warning, and comprehensive jam warning. The process of determining whether the traffic density coefficient of the monitored lane section is greater than the set first density coefficient threshold is as follows: When the pedestrian traffic density is greater than the set pedestrian traffic density threshold and the vehicle traffic density is less than the set vehicle traffic density threshold, determine whether the traffic density coefficient is greater than the set first pedestrian density coefficient threshold. If so, issue a pedestrian jam warning; when the pedestrian traffic density is less than the set pedestrian traffic density threshold and the vehicle traffic density is greater than the set vehicle traffic density threshold, determine whether the traffic density coefficient is greater than the set first vehicle density coefficient threshold. If so, issue a vehicle jam warning; when the pedestrian traffic density is greater than the set pedestrian traffic density threshold and the vehicle traffic density is greater than the set vehicle traffic density threshold, determine whether the traffic density coefficient is greater than the set first comprehensive density coefficient threshold. If so, issue a comprehensive jam warning.

[0034] In this implementation scheme, according to the situations of pedestrian traffic density and vehicle traffic density, the traffic jam warning is divided into pedestrian jam warning, vehicle jam warning, and comprehensive jam warning, making the warning more targeted and effective. By separately determining whether the pedestrian and vehicle traffic densities exceed the thresholds and issuing corresponding jam warnings, traffic congestion problems can be more accurately detected, making the warning more precise and effective.

[0035] Conducting different types of jam warnings according to different situations of pedestrian and vehicle densities is conducive to the traffic management department taking corresponding measures according to the actual situation, improving the efficiency and pertinence of traffic management.

[0036] Specifically, the second density coefficient threshold includes the second pedestrian density coefficient threshold, the second vehicle density coefficient threshold, and the second comprehensive density coefficient threshold. The traffic jam warning includes the pedestrian severe jam warning, the vehicle severe jam warning, and the comprehensive severe jam warning. The process of determining whether the traffic density coefficient in the monitored lane section is greater than the set second density coefficient threshold is as follows: When the pedestrian traffic density is greater than the set pedestrian traffic density threshold and the vehicle traffic density is less than the set vehicle traffic density threshold, determine whether the traffic density coefficient is greater than the set second pedestrian density coefficient threshold. If so, issue a pedestrian severe jam warning; when the pedestrian traffic density is less than the set pedestrian traffic density threshold and the vehicle traffic density is greater than the set vehicle traffic density threshold, determine whether the traffic density coefficient is greater than the set second vehicle density coefficient threshold. If so, issue a vehicle severe jam warning; when the pedestrian traffic density is greater than the set pedestrian traffic density threshold and the vehicle traffic density is greater than the set vehicle traffic density threshold, determine whether the traffic density coefficient is greater than the set second comprehensive density coefficient threshold. If so, issue a comprehensive severe jam warning.

[0037] First, according to the situations of pedestrian traffic density and vehicle traffic density, respectively determine whether different conditions are met. These conditions include whether the pedestrian traffic density is greater than the set pedestrian traffic density threshold, whether the vehicle traffic density is less than the set vehicle traffic density threshold, and the opposite situations. After determining that a specific condition is met, calculate the traffic density coefficient based on the obtained pedestrian traffic density and vehicle traffic density. This coefficient can be calculated in the manner given in the formula, considering the weight coefficients of pedestrian and vehicle densities, as well as the comparison with the threshold.

[0038] Compare the calculated traffic density coefficient with the set second density coefficient threshold. If the traffic density coefficient is greater than the set second density coefficient threshold, it indicates that the traffic density has exceeded the preset severe jam level, and corresponding severe jam warnings need to be issued. According to the judgment result, determine what type of severe jam warning to issue. If the condition met is that the pedestrian density exceeds the threshold while the vehicle density does not exceed the threshold, issue a pedestrian severe jam warning; otherwise, issue a vehicle severe jam warning. If both the pedestrian density and the vehicle density exceed the threshold simultaneously, issue a comprehensive severe jam warning.

[0039] By re - determining whether the pedestrian and vehicle traffic densities exceed the second density coefficient threshold and issuing corresponding severe jam warnings, traffic congestion problems can be detected more accurately, making the warning more precise and effective.

[0040] In addition to being directly represented by the update speed of pedestrians or vehicles within each determined period, the traffic update coefficient can be calculated in the following way. The calculation formula for the traffic update coefficient is:

[0041] ;

[0042] In the formula, is the traffic update coefficient, is the pedestrian update rate, is the set reference pedestrian update rate, is the vehicle update rate, is the set reference vehicle update rate, is the pedestrian update rate weight coefficient, is the vehicle update rate weight coefficient, and , is the natural constant, is the pedestrian traffic density, which is determined based on the number of pedestrians in the lane section to be monitored or the number of pedestrians in the adjacent monitored lane section, is the set pedestrian traffic density threshold, is the vehicle traffic density, is the set vehicle traffic density threshold.

[0043] In this implementation plan, the traffic update rate refers to the rate of change of the number of pedestrians and vehicles within a certain period, which can be represented by the update speed of pedestrians or vehicles in each determined period. The level of the traffic update rate can reflect the change of traffic flow, that is, the speed of change of the number of pedestrians and vehicles on the road. The calculation logic of the traffic update coefficient is based on the comparison and trade-off of the pedestrian and vehicle update rates. When the pedestrian or vehicle update rate is higher than the set reference rate, the traffic update coefficient will reflect a relatively high traffic update level. The calculation methods of the traffic update coefficient in different situations are different. According to different combinations of pedestrian and vehicle traffic densities, the corresponding calculation formulas are selected for calculation to more accurately reflect the change of traffic conditions.

[0044] The traffic update coefficient is based on the update rates of pedestrians and vehicles and can reflect the dynamic change of traffic conditions, enabling traffic management to make responses more timely. Considering the update rates of pedestrians and vehicles comprehensively, the traffic update coefficient can more comprehensively reflect the comprehensive situation of traffic conditions, which is beneficial to improving the comprehensiveness and accuracy of traffic management.

[0045] Specifically, the traffic update threshold coefficient includes a pedestrian update threshold coefficient, a vehicle update threshold coefficient, and a comprehensive update threshold coefficient. The process of determining whether the traffic update coefficient is less than the set traffic update threshold coefficient is as follows: When the pedestrian traffic density is greater than the set pedestrian traffic density threshold and the vehicle traffic density is less than the set vehicle traffic density threshold, determine whether the traffic update coefficient is less than the set pedestrian update threshold coefficient. If so, determine whether the traffic density coefficient in the adjacent monitored lane section is greater than the set second density coefficient threshold. If not, continue to determine whether the traffic update coefficient is less than the set pedestrian update threshold coefficient; When the pedestrian traffic density is less than the set pedestrian traffic density threshold and the vehicle traffic density is greater than the set vehicle traffic density threshold, determine whether the traffic update coefficient is less than the set vehicle update threshold coefficient. If so, determine whether the traffic density coefficient in the adjacent monitored lane section is greater than the set second density coefficient threshold. If not, continue to determine whether the traffic update coefficient is less than the set vehicle update threshold coefficient; When the pedestrian traffic density is greater than the set pedestrian traffic density threshold and the vehicle traffic density is greater than the set vehicle traffic density threshold, determine whether the traffic update coefficient is less than the set comprehensive update threshold coefficient. If so, determine whether the traffic density coefficient in the adjacent monitored lane section is greater than the set second density coefficient threshold. If not, continue to determine whether the traffic update coefficient is less than the set comprehensive update threshold coefficient.

[0046] In this implementation plan, first, according to the situations of pedestrian traffic density and vehicle traffic density, it is respectively determined whether specific conditions are met, including that the pedestrian density is greater than the set threshold and the vehicle density is less than the set threshold, the pedestrian density is less than the set threshold and the vehicle density is greater than the set threshold, and both are greater than the corresponding thresholds. According to different situations, corresponding processing and judgments are carried out.

[0047] Under the condition of meeting specific conditions, further determine whether the traffic update coefficient is less than the set traffic update threshold coefficient. The comparison of the traffic update coefficient is made by comparing the calculated traffic update coefficient with the set pedestrian update threshold coefficient, vehicle update threshold coefficient, or comprehensive update threshold coefficient to determine whether the warning condition is met. If the traffic update coefficient is less than the set traffic update threshold coefficient, continue with subsequent judgments, including determining whether the traffic density coefficient in the adjacent monitored lane section is greater than the set second density coefficient threshold, and continue to determine whether the traffic update coefficient is less than the set threshold coefficient. Through continuous judgments, the changing trend and severity of the traffic condition can be determined more accurately, providing more comprehensive data support for traffic management.

[0048] Specifically, the process of perceiving, monitoring, identifying, and tracking moving targets is as follows: Based on monitoring data, feature extraction is performed on moving targets entering the lane section to be monitored or the adjacent monitoring lane section, and a pedestrian feature matching dataset and a vehicle feature matching dataset are obtained and stored; Based on a set matching period, feature extraction is performed on moving targets in the lane section to be monitored or the adjacent monitoring lane section, a pedestrian feature dataset and a vehicle feature dataset are obtained, and the pedestrian feature dataset and the vehicle feature dataset are matched with the pedestrian feature matching dataset and the vehicle feature matching dataset; Based on the matching results, the moving targets are identified and tracked, and the moving pedestrians and moving vehicles in the lane section to be monitored or the adjacent monitoring lane section are determined. The moving pedestrians and moving vehicles in the lane section to be monitored serve as the analysis basis for the traffic density coefficient in the monitoring lane section, and the moving pedestrians and moving vehicles in the adjacent monitoring lane section serve as the analysis basis for the traffic density coefficient in the adjacent monitoring lane section.

[0049] In this implementation plan, first, based on monitoring data, feature extraction is performed on moving targets entering the lane section to be monitored or the adjacent monitoring lane section. The extracted feature data is stored in the pedestrian feature matching dataset and the vehicle feature matching dataset for subsequent matching and identification processes. During the set matching period, feature extraction is performed again on moving targets in the lane section to be monitored or the adjacent monitoring lane section, and a pedestrian feature dataset and a vehicle feature dataset are obtained. These newly extracted feature data are matched with the previously stored pedestrian feature matching dataset and vehicle feature matching dataset to find the corresponding moving targets.

[0050] Based on the matching results, the moving targets are identified and tracked. Through the successfully matched feature data, the moving pedestrians and moving vehicles in the lane section to be monitored or the adjacent monitoring lane section are determined. These identified and tracked moving targets will serve as the basis for subsequent traffic density coefficient analysis.

[0051] Through periodic feature extraction and matching, real-time perception and monitoring of moving targets can be achieved, timely reflecting the changes in traffic conditions in the lane section to be monitored and the adjacent monitoring lane section. Moreover, through feature extraction and matching, moving targets can be accurately identified and tracked, reducing the possibility of misidentification and missed identification, and improving the accuracy of identification and tracking. By performing feature extraction and matching on pedestrians and vehicles, perception and monitoring of different types of moving targets are realized, thus comprehensively understanding the traffic conditions and providing more comprehensive data support for subsequent traffic density coefficient analysis.

[0052] Specifically, the pedestrian feature dataset includes a facial feature dataset, a head feature dataset, and a clothing feature dataset. The pedestrian feature matching dataset includes a facial feature matching dataset, a head feature matching dataset, and a clothing feature matching dataset. The pedestrian feature dataset and the pedestrian feature matching dataset are matched to obtain a pedestrian matching coefficient, which serves as the basis for identifying and tracking moving targets and also as the analysis basis for traffic update coefficients. The vehicle feature dataset and the vehicle feature matching dataset are matched to obtain a vehicle matching coefficient, which serves as the basis for identifying and tracking moving targets and also as the analysis basis for traffic update coefficients.

[0053] The pedestrian feature dataset includes a facial feature dataset, a head feature dataset, and a clothing feature dataset, while the pedestrian feature matching dataset includes corresponding matching datasets for storing previously acquired feature data. Similarly, the vehicle feature dataset and the vehicle feature matching dataset also have a similar structure for storing vehicle feature data and matching data.

[0054] By matching the pedestrian feature dataset with the pedestrian feature matching dataset, a pedestrian matching coefficient can be obtained, which reflects the matching degree between the pedestrian feature data and the previously stored feature data. Similarly, the vehicle feature dataset is matched with the vehicle feature matching dataset to obtain a vehicle matching coefficient.

[0055] The facial feature matching dataset includes the area of facial occluders, the curvature of occluders, and the bending angle of occluders. The head feature dataset includes the distance between eyebrows, the width of the lips, and the distance from the upper lip to the eyebrows. The clothing feature dataset includes the width value of the upper garment and the curvature of the neckline. The facial feature matching dataset includes the matching area of facial occluders, the matching curvature of occluders, and the matching bending angle of occluders. The head feature matching dataset includes the matching distance between eyebrows, the matching lip width, and the matching distance from the upper lip to the eyebrows. The clothing feature matching dataset includes the matching width value of the upper garment and the matching curvature of the neckline. If the area of the facial occluder is less than the set threshold of the facial occluder area, the pedestrian matching coefficient is analyzed based on the head feature dataset, the clothing feature dataset, the head feature matching dataset, and the clothing feature matching dataset. If the area of the facial occluder is greater than the set threshold of the facial occluder area, the pedestrian matching coefficient is analyzed based on the facial feature dataset, the clothing feature dataset, the facial feature matching dataset, and the clothing feature matching dataset.

[0056] According to the descriptions of the facial feature dataset, head feature dataset, and clothing feature dataset, facial, head, and clothing feature data related to the human body are extracted. These feature data are matched with the corresponding feature matching datasets for subsequent analysis of the matching degree of pedestrians. The area of the facial occluder is an important feature for judging whether the face is occluded. If the area of the facial occluder is less than the set threshold, the head features and clothing features are used for matching; if the area of the facial occluder is greater than the set threshold, the facial features and clothing features are used for matching. According to different matching situations, the head feature dataset and clothing feature dataset or the facial feature dataset and clothing feature dataset are respectively used, combined with the corresponding matching datasets, for matching analysis.

[0057] The influence of the facial occluder on matching is considered. It is not limited to the matching of facial features only, but also takes into account head features and clothing features, making the matching analysis more comprehensive. Different matching schemes are adjusted according to the facial occlusion situation, which is more flexible and adaptable, and can meet the human body feature matching requirements in various different scenarios.

[0058] In addition to being obtained by deep learning algorithms or Siamese networks based on deep learning for pedestrian matching, the pedestrian matching coefficient can also be obtained through the following calculation formula:

[0059] ;

[0060] In the formula, is the number of the pedestrian feature matching dataset, , is the total number of the pedestrian feature matching datasets, is the pedestrian matching coefficient obtained by matching the pedestrian feature dataset with the th pedestrian feature matching dataset, is the head feature dataset, is the head feature matching dataset, is the calculation function for matching the head feature dataset with the head feature matching dataset in the th pedestrian feature matching dataset, is the weight factor of the calculation function , is the clothing feature dataset, is the clothing feature matching dataset, is the calculation function for matching the clothing feature dataset with the clothing feature matching dataset in the th pedestrian feature matching dataset, is the weight factor of the calculation function , is the facial feature dataset, is a facial feature matching dataset, is a calculation function for matching the facial feature dataset with the facial feature matching dataset in the th pedestrian feature matching dataset, is the weight factor of the calculation function ; is the area of the facial occluder, is the set threshold of the facial occluder area.

[0061] The calculation formula of the matching coefficient uses multiple feature datasets (head, clothing, face) and their matching functions with the matching dataset. The formula is divided into two cases: when the area of the facial occluder is less than the set threshold, the head features and clothing features are used for matching; when the area of the facial occluder is greater than the set threshold, the head features and face features are used for matching. The matching function is used to evaluate the matching degree between the feature dataset and the matching dataset. These functions can calculate the similarity between features based on various similarity measurement methods, such as Euclidean distance, cosine similarity, etc., or can be calculated through the following calculation methods.

[0062] Among them, , is the distance between the eyebrows, is the th eyebrow matching distance in the head feature matching dataset of the th pedestrian feature matching dataset, is the width of the lips, is the matching lip width in the head feature matching dataset of the th pedestrian feature matching dataset, is the distance from the upper lip to the eyebrows, is the matching distance from the upper lip to the eyebrows in the head feature matching dataset of the th pedestrian feature matching dataset, is the comprehensive weight factor of the distance between the eyebrows and the width of the lips,

[0063] ,

[0064] In the formula, is the value of the width of the upper garment, is the th matching width value of the upper garment in the clothing feature matching data of the th pedestrian feature matching dataset, is the weight factor of the value of the width of the upper garment, is the curvature of the neckline, is the matching curvature of the neckline in the clothing feature matching data of the th pedestrian feature matching dataset, is the weight factor of the curvature of the neckline.

[0065] ,

[0066] Wherein, is the area of the facial occluder, is the th facial occluder matching area in the facial feature matching dataset of the th pedestrian feature matching dataset, is the weight factor of the facial occluder area, is the curvature of the occluder, is the th occluder matching curvature in the facial feature matching dataset of the th pedestrian feature matching dataset, is the occluder matching curvature angle, is the curvature weight factor.

[0067] For each matching parameter in the above pedestrian feature matching dataset, it is obtained by acquiring the features of each pedestrian in the current cycle. After extracting the features of each pedestrian, they are stored in the database, and the matching parameters stored in the previous monitoring cycle are deleted in the next monitoring cycle to achieve update.

[0068] The vehicle feature dataset includes the windshield area and the vehicle HSV value, and the vehicle feature matching dataset includes the windshield matching area and the vehicle HSV matching value.

[0069] The vehicle matching coefficient can be obtained not only through deep learning algorithms or Siamese networks based on deep learning, but also through the following calculation formula:

[0070] ;

[0071] Wherein, is the number of the vehicle feature matching dataset, , is the total number of the vehicle feature matching dataset, is the windshield area, is the th windshield matching area in the th vehicle feature matching dataset, is the windshield area weight factor, is the H value in the vehicle HSV value, is the H value of the vehicle HSV matching value in the th vehicle feature matching dataset, is the The S value of the vehicle HSV matching value in a vehicle feature matching dataset, is the V value in the vehicle HSV value, is the V value of the vehicle HSV matching value in the nth vehicle feature matching dataset, and

[0072] is the HSV value weight factor. Similarly, for each matching parameter in the above vehicle feature matching dataset, it is obtained by acquiring the features of each vehicle in the current period. After extracting the features of each vehicle, they are stored in the database, and the matching parameters stored in the previous monitoring period are deleted in the next monitoring period to achieve update.

[0073] In this implementation plan, the calculation of the vehicle matching coefficient takes into account two aspects: the windshield area and the vehicle HSV value, making the matching process more comprehensive, enabling the vehicle to be identified and tracked from multiple angles. The weight factor can be adjusted according to the actual situation, making the algorithm have a certain degree of flexibility and adaptability, and can be adjusted according to different scenarios and requirements, improving the applicability of the algorithm.

[0074] A traffic management auxiliary perception method and device based on radar image information fusion, comprising a moving target perception and monitoring module, a monitoring and warning module for the lane section to be monitored, a traffic update monitoring module for the lane section to be monitored, and a monitoring and warning module for the adjacent lane section to be monitored, wherein: The moving target perception and monitoring module is used to fuse the monitoring data with the radar image information, perceive and monitor the moving targets in the lane section to be monitored, and identify and track the moving targets in the lane section to be monitored; The monitoring and warning module for the lane section to be monitored is used to determine the traffic density coefficient of the lane section to be monitored in the current monitoring period based on the identified and tracked moving targets in the lane section to be monitored, and judge whether the traffic density coefficient of the monitored lane section in the current monitoring period is greater than the set first density coefficient threshold. If not, continue to judge whether the traffic density coefficient of the monitored lane section in the next monitoring period is greater than the set first density coefficient threshold. If so, issue a traffic jam warning; The traffic update monitoring module for the lane section to be monitored is used to obtain the traffic update coefficient of the lane section to be monitored in the current monitoring period, and judge whether the traffic update coefficient in the current monitoring period is less than the set traffic update threshold coefficient. If so, perform perception and monitoring on the adjacent lane section to be monitored. If not, when the traffic density coefficient of the lane section to be monitored in the next monitoring period is greater than the set first density coefficient threshold, judge whether the traffic update coefficient in the next monitoring period is less than the set traffic update threshold coefficient; The monitoring and warning module for the adjacent lane section to be monitored is used to identify and track the moving targets in the adjacent lane section to be monitored based on the perception and monitoring of the adjacent lane section to be monitored, determine the traffic density coefficient of the adjacent lane section to be monitored in the current monitoring period, and judge whether the traffic density coefficient of the adjacent lane section to be monitored in the current monitoring period is greater than the set second density coefficient threshold. If so, issue a first-level severe traffic jam warning. If not, judge whether the traffic update coefficient of the adjacent lane section to be monitored in the current monitoring period is less than the set traffic update threshold coefficient. If so, issue a second-level severe traffic jam warning. If not, when the traffic density coefficient of the lane section to be monitored in the next monitoring period is greater than the set first density coefficient threshold and the traffic density coefficient of the adjacent lane section to be monitored in the next monitoring period is greater than the set second density coefficient threshold, judge whether the traffic update coefficient of the adjacent lane section to be monitored in the next monitoring period is less than the set traffic update threshold coefficient.

[0075] In this implementation plan, through the moving target perception and monitoring module, the monitoring data is fused with the radar image information to achieve comprehensive perception, identification, and tracking of moving targets, improving the accuracy and comprehensiveness of traffic conditions. The monitoring and warning module for the lane section to be monitored determines the traffic density coefficient of the lane section to be monitored based on the identified and tracked moving targets, and continuously judges whether it exceeds the set first density coefficient threshold. If it exceeds the threshold, traffic updates are performed and a traffic jam warning is issued, thus timely reminding drivers and traffic managers to take measures.

[0076] The traffic update monitoring module for the lane section to be monitored obtains the traffic update coefficient and determines whether it is less than the set traffic update threshold coefficient. If it is less than the threshold, adjacent lane section perception monitoring is performed, realizing continuous monitoring and update of the traffic condition. The adjacent lane section monitoring and warning module determines the traffic density coefficient of the adjacent lane section according to the identified and tracked moving targets, and determines whether it exceeds the set second density coefficient threshold. If it does not exceed the threshold, it further determines whether the traffic update coefficient of the lane section to be monitored is less than the set traffic update threshold coefficient. If so, a traffic severe congestion warning is issued, improving the ability to detect and handle traffic anomalies in a timely manner.

[0077] An electronic device includes: a processor; and a memory in which computer program instructions are stored. When the computer program instructions are run by the processor, the processor executes the above traffic management auxiliary perception method based on radar image information fusion.

[0078] A computer-readable storage medium is used to store a program. When the program is executed by a processor, the above traffic management auxiliary perception method based on radar image information fusion is implemented.

[0079] In the embodiments of the present invention, by integrating monitoring data and radar image information, and performing perception monitoring, identification and tracking on moving targets, comprehensive perception and tracking of traffic scenarios are realized, thereby effectively monitoring traffic conditions. Using information such as the density of pedestrians and vehicles, combined with relevant calculation formulas and thresholds, real-time analysis and warning of traffic density are realized, and problems such as traffic congestion can be discovered in a timely manner.

[0080] By setting different density coefficient thresholds and update thresholds, and comprehensively considering the matching results of different features, accurate warning of severe congestion of pedestrians and vehicles is realized, which helps to take traffic management measures in a timely manner. Multi-dimensional feature matching is adopted, such as the facial features, head features, clothing features of pedestrians, and the windshield area and HSV value of vehicles, etc., improving the accuracy and reliability of matching.

[0081] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) containing computer-usable program codes.

[0082] The present invention is described with reference to flowchart illustrations and / or block diagrams of systems, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each flow and / or block in the flowchart illustrations and / or block diagrams, and combinations of flows and / or blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions executed by the processor of the computer or other programmable data processing apparatus create means for implementing the functions specified in the flow Figure 1 one or more flows and / or blocks Figure 1 one or more blocks.

[0083] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means that implement the functions specified in the flow Figure 1 one or more flows and / or blocks Figure 1 one or more blocks.

[0084] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, such that the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in the flow Figure 1 one or more flows and / or blocks Figure 1 one or more blocks.

[0085] Although the preferred embodiments of the present invention have been described, additional changes and modifications can be made by those skilled in the art once they learn of the basic inventive concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0086] It is apparent that those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims

1. A traffic management auxiliary perception method based on radar image information fusion, characterized in that: The following steps are involved: The monitoring data is integrated with the radar image information to sense and monitor the moving targets in the lane section to be monitored, and the moving targets in the lane section to be monitored are identified and tracked; Determine the traffic density coefficient of the lane section to be monitored in the current monitoring cycle based on the identified and tracked moving target in the lane section to be monitored, and judge whether the traffic density coefficient of the lane section to be monitored in the current monitoring cycle is greater than the set first density coefficient threshold value, if not, continue to judge whether the traffic density coefficient of the lane section to be monitored in the next monitoring cycle is greater than the set first density coefficient threshold value, if yes, issue a traffic jam warning; Obtain the traffic update coefficient of the lane section to be monitored in the current monitoring cycle, and determine whether the traffic update coefficient of the current monitoring cycle is less than the set traffic update threshold coefficient. If so, perform perception monitoring of the adjacent monitoring lane section. Otherwise, when the traffic density coefficient of the lane section to be monitored in the next monitoring cycle is greater than the set first density coefficient threshold, determine whether the traffic update coefficient of the next monitoring cycle is less than the set traffic update threshold coefficient. Based on the perception monitoring of the adjacent monitoring lane interval, the moving target of the adjacent monitoring lane interval is identified and tracked, and the traffic density coefficient of the adjacent monitoring lane interval of the current monitoring cycle is determined, and it is judged whether the traffic density coefficient of the adjacent monitoring lane interval of the current monitoring cycle is greater than the set second density coefficient threshold. If so, a first-level serious warning of traffic congestion is issued, and it is judged whether the traffic update coefficient of the adjacent monitoring lane interval of the current monitoring cycle is less than the set traffic update threshold coefficient. If so, a second-level serious warning of traffic congestion is issued. Otherwise, when the traffic density coefficient of the lane interval to be monitored in the next monitoring cycle is greater than the set first density coefficient threshold and the traffic density coefficient of the adjacent monitoring lane interval in the next monitoring cycle is greater than the set second density coefficient threshold, it is judged whether the traffic update coefficient of the adjacent monitoring lane interval in the next monitoring cycle is less than the set traffic update threshold coefficient; The moving targets include moving pedestrians and moving vehicles, and the calculation formula of the traffic density coefficient is: Wherein, CmX is the traffic density coefficient, Xr is the pedestrian traffic density, the pedestrian traffic density is determined based on the number of pedestrians in the lane section to be monitored or the number of pedestrians in the lane section adjacent to the monitoring lane, Xrb is the set pedestrian traffic density threshold, Cr is the vehicle traffic density, Crb is the set vehicle traffic density threshold, α1 is the pedestrian density weight coefficient, α2 is the vehicle density weight coefficient, and α1+α2=1; The calculation formula of the traffic update coefficient is: Wherein, GgX is the traffic update coefficient, Xg is the pedestrian update rate, Xgb is the set pedestrian update reference rate, Cg is the vehicle update rate, Cgb is the set vehicle update reference rate, β1 is the pedestrian update rate weight coefficient, β2 is the vehicle update rate weight coefficient, and β1+β2=1, e is a natural constant, Xr is the pedestrian traffic density, which is determined based on the number of pedestrians in the lane section to be monitored or the number of pedestrians in the adjacent monitoring lane section, Xrb is the set pedestrian traffic density threshold, Cr is the vehicle traffic density, and Crb is the set vehicle traffic density threshold.

2. The traffic management auxiliary perception method based on radar image information fusion according to claim 1 is characterized in that The first density coefficient threshold includes a first pedestrian density coefficient threshold, a first vehicle density coefficient threshold and a first comprehensive density coefficient threshold. The traffic congestion warning includes a pedestrian congestion warning, a vehicle congestion warning and a comprehensive congestion warning. The process of judging whether the traffic density coefficient of the lane section to be monitored is greater than the set first density coefficient threshold is as follows: When the pedestrian traffic density is greater than the set pedestrian traffic density threshold and the vehicle traffic density is less than the set vehicle traffic density threshold, it is determined whether the traffic density coefficient is greater than the set first pedestrian density coefficient threshold. If so, a pedestrian congestion warning is issued; When the pedestrian traffic density is less than the set pedestrian traffic density threshold and the vehicle traffic density is greater than the set vehicle traffic density threshold, it is determined whether the traffic density coefficient is greater than the set first vehicle density coefficient threshold, and if so, a vehicle congestion warning is issued; When the pedestrian traffic density is greater than the set pedestrian traffic density threshold and the vehicle traffic density is greater than the set vehicle traffic density threshold, it is determined whether the traffic density coefficient is greater than the set first comprehensive density coefficient threshold. If so, a comprehensive congestion warning is issued.

3. The traffic management auxiliary perception method based on radar image information fusion according to claim 1 is characterized in that: The second density coefficient threshold includes a second pedestrian density coefficient threshold, a second vehicle density coefficient threshold and a second comprehensive density coefficient threshold. The traffic congestion warning includes a pedestrian severe congestion warning, a vehicle severe congestion warning and a comprehensive severe congestion warning. The process of judging whether the traffic density coefficient of the adjacent monitoring lane section is greater than the set second density coefficient threshold is as follows: When the pedestrian traffic density is greater than the set pedestrian traffic density threshold and the vehicle traffic density is less than the set vehicle traffic density threshold, it is determined whether the traffic density coefficient is greater than the set second pedestrian density coefficient threshold. If so, a serious pedestrian congestion warning is issued; When the pedestrian traffic density is less than the set pedestrian traffic density threshold and the vehicle traffic density is greater than the set vehicle traffic density threshold, it is determined whether the traffic density coefficient is greater than the set second vehicle density coefficient threshold. If so, a serious vehicle congestion warning is issued; When the pedestrian traffic density is greater than the set pedestrian traffic density threshold and the vehicle traffic density is greater than the set vehicle traffic density threshold, it is determined whether the traffic density coefficient is greater than the set second comprehensive density coefficient threshold. If so, a comprehensive severe congestion warning is issued.

4. The traffic management auxiliary perception method based on radar image information fusion according to claim 1 is characterized in that: The traffic update threshold coefficient includes a pedestrian update threshold coefficient, a vehicle update threshold coefficient and a comprehensive update threshold coefficient. The process of judging whether the traffic update coefficient is less than the set traffic update threshold coefficient is as follows: When the pedestrian traffic density is greater than the set pedestrian traffic density threshold and the vehicle traffic density is less than the set vehicle traffic density threshold, determine whether the traffic update coefficient is less than the set pedestrian update threshold coefficient. If so, determine whether the traffic density coefficient of the adjacent monitoring lane interval is greater than the set second density coefficient threshold. If not, continue to determine whether the traffic update coefficient is less than the set pedestrian update threshold coefficient. When the pedestrian traffic density is less than the set pedestrian traffic density threshold and the vehicle traffic density is greater than the set vehicle traffic density threshold, determine whether the traffic update coefficient is less than the set vehicle update threshold coefficient. If so, determine whether the traffic density coefficient of the adjacent monitoring lane interval is greater than the set second density coefficient threshold. If not, continue to determine whether the traffic update coefficient is less than the set vehicle update threshold coefficient. When the pedestrian traffic density is greater than the set pedestrian traffic density threshold and the vehicle traffic density is greater than the set vehicle traffic density threshold, determine whether the traffic update coefficient is less than the set comprehensive update threshold coefficient. If so, determine whether the traffic density coefficient of the adjacent monitoring lane interval is greater than the set second density coefficient threshold. If not, continue to determine whether the traffic update coefficient is less than the set comprehensive update threshold coefficient.

5. The traffic management auxiliary perception method based on radar image information fusion according to claim 1 is characterized in that: The process of sensing, monitoring, identifying and tracking moving targets is as follows: Based on the monitoring data, feature extraction is performed on the moving targets that enter the lane section to be monitored or the lane section adjacent to the monitored lane section, and a pedestrian feature matching data set and a vehicle feature matching data set are obtained and stored; Based on the set matching cycle, feature extraction is performed on the moving target in the monitored lane section or the adjacent monitored lane section, a pedestrian feature data set and a vehicle feature data set are obtained, and the pedestrian feature data set and the vehicle feature data set are matched with the pedestrian feature matching data set and the vehicle feature matching data set; Based on the matching results, the moving targets are identified and tracked, and the moving pedestrians and moving vehicles in the lane section to be monitored or the adjacent monitoring lane section are determined. The moving pedestrians and moving vehicles in the lane section to be monitored are used as the analysis basis for the traffic density coefficient of the monitoring lane section, and the moving pedestrians and moving vehicles in the adjacent monitoring lane section are used as the analysis basis for the traffic density coefficient of the adjacent monitoring lane section.

6. The traffic management auxiliary perception method based on radar image information fusion according to claim 5 is characterized by: The pedestrian feature data set includes a facial feature data set, a head feature data set and a clothing feature data set, the pedestrian feature matching data set includes a facial feature matching data set, a head feature matching data set and a clothing feature matching data set, the pedestrian feature data set and the pedestrian feature matching data set are matched to obtain a pedestrian matching coefficient, the pedestrian matching coefficient is used as a basis for identifying and tracking moving targets, and is also used as a basis for analyzing traffic update coefficients; The vehicle feature data set and the vehicle feature matching data set are matched to obtain a vehicle matching coefficient, which is used as a basis for identifying and tracking moving targets and also as a basis for analyzing traffic update coefficients.

7. The traffic management auxiliary perception method based on radar image information fusion according to claim 6 is characterized in that: The facial feature matching data set includes the facial occluder area, occluder curvature and occluder curvature angle, the head feature data set includes the distance between eyebrows, lip width and the distance from upper lip to eyebrows, and the clothing feature data set includes the width value of tops and the curvature of the neckline; The facial feature matching data set includes facial occluder matching area, occluder matching curvature and occluder matching curvature angle; the head feature matching data set includes glabellar matching distance, matching lip width and upper lip to glabellar matching distance; the clothing feature matching data set includes top matching width value and neckline matching curvature; If the facial occlusion area is smaller than the set facial occlusion area threshold, the pedestrian matching coefficient is obtained based on the head feature data set and the clothing feature data set as well as the head feature matching data set and the clothing feature matching data set. If the facial occlusion area is larger than the set facial occlusion area threshold, the pedestrian matching coefficient is obtained based on the facial feature data set and the clothing feature data set as well as the facial feature matching data set and the clothing feature matching data set. The vehicle feature data set includes a windshield area and a vehicle HSV value, and the vehicle feature matching data set includes a windshield matching area and a vehicle HSV matching value.

8. A device for applying a traffic management auxiliary perception method based on radar image information fusion as described in any one of claims 1 to 7, characterized in that: It includes a mobile target perception monitoring module, a monitoring and early warning module for the lane interval to be monitored, a traffic update monitoring module for the lane interval to be monitored, and a monitoring and early warning module for the adjacent monitoring lane interval, among which: The mobile target sensing and monitoring module is used to fuse the monitoring data with the radar image information, sense and monitor the mobile targets in the lane section to be monitored, and identify and track the mobile targets in the lane section to be monitored; The monitoring and early warning module for the lane interval to be monitored is used to determine the traffic density coefficient of the lane interval to be monitored in the current monitoring cycle based on the mobile target of the lane interval to be monitored that is identified and tracked, and to judge whether the traffic density coefficient of the lane interval to be monitored in the monitoring cycle is greater than the set first density coefficient threshold value, if not, to continue to judge whether the traffic density coefficient of the lane interval to be monitored in the next monitoring cycle is greater than the set first density coefficient threshold value, and if so, to issue a traffic jam early warning; The traffic update monitoring module for the lane interval to be monitored is used to obtain the traffic update coefficient of the lane interval to be monitored in the current monitoring cycle, and to determine whether the traffic update coefficient of the current monitoring cycle is less than the set traffic update threshold coefficient. If so, perception monitoring of the adjacent monitoring lane interval is performed. Otherwise, when the traffic density coefficient of the lane interval to be monitored in the next monitoring cycle is greater than the set first density coefficient threshold, it is determined whether the traffic update coefficient of the next monitoring cycle is less than the set traffic update threshold coefficient. The adjacent monitoring lane interval monitoring and early warning module is used to identify and track moving targets in the adjacent monitoring lane interval based on the perception monitoring of the adjacent monitoring lane interval, and determine the traffic density coefficient of the adjacent monitoring lane interval in the current monitoring cycle, and judge whether the traffic density coefficient of the adjacent monitoring lane interval in the current monitoring cycle is greater than the set second density coefficient threshold. If so, a first-level severe warning of traffic congestion is issued, and it is judged whether the traffic update coefficient of the adjacent monitoring lane interval in the current monitoring cycle is less than the set traffic update threshold coefficient. If so, a second-level severe warning of traffic congestion is issued. Otherwise, when the traffic density coefficient of the lane interval to be monitored in the next monitoring cycle is greater than the set first density coefficient threshold and the traffic density coefficient of the adjacent monitoring lane interval in the next monitoring cycle is greater than the set second density coefficient threshold, it is judged whether the traffic update coefficient of the adjacent monitoring lane interval in the next monitoring cycle is less than the set traffic update threshold coefficient.

Images