Pedestrian safety early warning method and device combined with high-precision sensing technology

Abstract

The invention provides a pedestrian safety early warning method and device in combination with a high-precision perception technology, and relates to the technical field of intelligent transportation, and the method comprises the steps: positioning a target early warning region; presetting an early warning intervention window, and carrying out environment intermittent perception; time sequence risk feature identification is carried out based on the information perception time sequence; performing risk trajectory prediction on the dynamic risk factor distribution to obtain early warning intervention risk distribution; performing risk level evaluation according to the early warning intervention risk distribution, and outputting a real-time risk level; matching a traffic flow collaborative early warning strategy; and taking the target early-warning area as an early-warning boundary, and executing traffic flow collaborative early-warning optimization on pedestrian safety by adopting a traffic flow collaborative early-warning strategy. According to the method and the device, the technical problem of poor early warning effect caused by difficulty in accurately identifying pedestrian emergency behaviors due to perception limitation is solved, early warning intervention is performed in advance through time sequence risk feature identification in combination with trajectory prediction, the accuracy of pedestrian safety early warning is improved, and the optimization of the whole traffic flow is realized.

Description

Technical Field

[0001] This application relates to the field of intelligent transportation technology, and in particular, to a pedestrian safety warning method and device combining high-precision sensing technology. Background Art

[0002] With the continuous development of urban transportation systems, pedestrian safety issues have received increasing attention. As a vulnerable group in the transportation system, pedestrians are easily threatened by motor vehicles in complex traffic environments. To reduce traffic accidents and improve road traffic safety, pedestrian safety warning systems have gradually become an important direction in intelligent transportation research. Pedestrian safety warning methods mainly rely on technical means such as visual perception, radar detection, and vehicle-road cooperation, and achieve early warning through real-time perception of pedestrian positions and environmental factors.

[0003] However, existing methods usually rely on fixed sensing points, making it difficult to identify sudden pedestrian behaviors in advance, which affects the timeliness and accuracy of early warnings. Due to the lack of sufficient combination of temporal risk feature analysis, the pedestrian warning system has a high false alarm rate and serious missed detections, affecting the response efficiency of drivers and pedestrians.

[0004] In summary, there are technical problems in the prior art that due to sensing limitations, it is difficult to accurately identify sudden pedestrian behaviors, resulting in poor warning effects and thus affecting overall traffic safety. Summary of the Invention

[0005] The purpose of this application is to provide a pedestrian safety warning method and device combining high-precision sensing technology to solve the technical problems in the prior art that due to sensing limitations, it is difficult to accurately identify sudden pedestrian behaviors, resulting in poor warning effects and thus affecting overall traffic safety.

[0006] In view of the above problems, this application provides a pedestrian safety warning method and device combining high-precision sensing technology.

[0007] In a first aspect, the present application provides a pedestrian safety warning method combined with high-precision perception technology. The pedestrian safety warning method combined with high-precision perception technology is implemented through a pedestrian safety warning device combined with high-precision perception technology. Among them, the pedestrian safety warning method combined with high-precision perception technology includes: through warning intervention analysis of the target risk area, locating the target warning area; presetting a warning intervention window, and based on the warning intervention window, performing intermittent environmental perception of the target warning area to obtain multiple regional dynamic information sets; based on the information perception time sequence, performing time sequence risk feature recognition of the multiple regional dynamic information sets to locate the distribution of dynamic risk factors; using the target risk area as the risk boundary, performing risk trajectory prediction on the distribution of dynamic risk factors to obtain a warning intervention risk distribution; performing risk level evaluation according to the warning intervention risk distribution, and outputting a real-time risk level; performing warning strategy matching according to the real-time risk level, and outputting a traffic flow collaborative warning strategy; using the target warning area as the warning boundary, and using the traffic flow collaborative warning strategy to perform traffic flow collaborative warning optimization on pedestrian safety.

[0008] Optionally, based on the target risk area, network data is called to obtain driving speed characteristics; the warning intervention response duration is obtained through interaction; according to the warning intervention response duration and the acceleration speed characteristics, the warning intervention response distance is calculated; starting from the target risk area, with the warning intervention response distance as the road network selection scale, the external road network is selected to obtain the target warning area.

[0009] Optionally, based on the target warning area, network data is called to obtain the external road network traffic flow characteristics and the external road network pedestrian flow characteristics; the first perception configuration strategy is obtained by comparing the external road network traffic flow characteristics with the perception function number table; the second perception configuration strategy is obtained by comparing the external road network pedestrian flow characteristics with the perception function number table; by fusing the first perception configuration strategy and the second perception configuration strategy, the target perception configuration strategy is output; the high-precision perception array of the target warning area is configured with the target perception configuration strategy as a constraint, where the high-precision perception array includes multiple perception nodes configured in multiple perception areas.

[0010] Optionally, by fusing and analyzing the external road network traffic flow characteristics and the external road network pedestrian flow characteristics, the effective duration of warning intervention is obtained; based on the effective duration of warning intervention, the localization of the warning intervention window is performed; with the warning intervention window as a constraint, the synchronous intermittent operation of multiple perception nodes in the high-precision perception array is performed to obtain multiple regional dynamic perception sequences of the multiple perception areas; after performing time sequence fusion on the multiple regional dynamic perception sequences, risk feature detection is performed based on the fusion result to obtain the multiple regional dynamic information sets.

[0011] Optionally, perform spatio-temporal synchronization on the multiple regional dynamic perception sequences according to the early warning intervention window to obtain multiple groups of regional dynamic perception sequences; perform multi-view image sequence synchronous splicing on the multiple groups of regional dynamic perception sequences according to the mapping relationship between the multiple perception regions and the target early warning region, and output multiple regional spliced perception images; perform vehicle flow feature detection on the first regional spliced perception image to obtain K vehicle dynamic features of K real-time vehicles, where the vehicle dynamic features include vehicle identification features, driving speed features, and driving direction features; perform pedestrian flow feature detection on the first regional spliced perception image to obtain F pedestrian dynamic features of F real-time pedestrians, where the pedestrian dynamic features include pedestrian identification features, walking speed features, and walking direction features; use the first perception time of the first regional spliced perception image to perform associated storage of the K vehicle dynamic features and the F pedestrian dynamic features to obtain a first regional dynamic information set; and so on, perform risk feature detection on the multiple regional spliced perception images to obtain the multiple regional dynamic information sets.

[0012] Optionally, based on the consistency of vehicle identification features, perform temporal aggregation of driving change features on the multiple regional dynamic information sets to obtain K first feature change sequences of the K real-time vehicles; based on the consistency of pedestrian identification features, perform temporal aggregation of walking change features on the multiple regional dynamic information sets to obtain F second feature change sequences of the F real-time pedestrians; preset driving feature change scales, where the driving feature change scales include driving speed change scales and driving direction change scales; traverse the K first feature change sequences using the driving feature change scales to perform risk judgment on the K real-time vehicles, and output M risk vehicles; and so on, preset walking feature change scales to perform risk judgment on the F real-time pedestrians, and output N risk pedestrians; extract M vehicle position features of the M risk vehicles and N pedestrian position features of the N risk pedestrians as the dynamic risk factor distribution.

[0013] Optionally, extract the M first feature change sequences of the M risk vehicles; extract the N second feature change sequences of the N risk pedestrians; use the target risk region as the risk boundary and the target early warning region as the warning boundary, and perform driving prediction on the M risk vehicles according to the M first feature change sequences to obtain a vehicle intervention risk distribution; use the target risk region as the risk boundary and the target early warning region as the warning boundary, and perform walking prediction on the N risk pedestrians according to the N second feature change sequences to obtain a pedestrian intervention risk distribution; perform cross-fusion of the vehicle intervention risk distribution and the pedestrian intervention risk distribution based on the risk position distribution, and output the early warning intervention risk distribution.

[0014] In a second aspect, the present application also provides a pedestrian safety warning device incorporating high-precision sensing technology, which is used to execute the pedestrian safety warning method incorporating high-precision sensing technology as described in the first aspect. Among them, the pedestrian safety warning device incorporating high-precision sensing technology includes: a warning intervention analysis module, which is used to locate a target warning area by performing warning intervention analysis on a target risk area; an environmental intermittent sensing module, which is used to preset a warning intervention window and perform environmental intermittent sensing of the target warning area based on the warning intervention window to obtain a plurality of regional dynamic information sets; a risk factor location module, which is used to identify the temporal risk characteristics of the plurality of regional dynamic information sets based on the information sensing time sequence and locate the distribution of dynamic risk factors; a risk trajectory prediction module, which is used to perform risk trajectory prediction on the distribution of the dynamic risk factors with the target risk area as the risk boundary to obtain a warning intervention risk distribution; a risk level determination module, which is used to perform risk level evaluation according to the warning intervention risk distribution and output a real-time risk level; a warning strategy matching module, which is used to match warning strategies according to the real-time risk level and output a traffic flow collaborative warning strategy; a collaborative warning module, which is used to perform traffic flow collaborative warning optimization on pedestrian safety by using the traffic flow collaborative warning strategy with the target warning area as the warning boundary.

[0015] One or more technical solutions provided in the present application have at least the following beneficial effects:

[0016] By performing warning intervention analysis on the target risk area to locate the target warning area; presetting a warning intervention window and performing environmental intermittent sensing of the target warning area based on the warning intervention window to obtain a plurality of regional dynamic information sets; identifying the temporal risk characteristics of the plurality of regional dynamic information sets based on the information sensing time sequence and locating the distribution of dynamic risk factors; performing risk trajectory prediction on the distribution of the dynamic risk factors with the target risk area as the risk boundary to obtain a warning intervention risk distribution; performing risk level evaluation according to the warning intervention risk distribution and outputting a real-time risk level; matching warning strategies according to the real-time risk level and outputting a traffic flow collaborative warning strategy; and performing traffic flow collaborative warning optimization on pedestrian safety by using the traffic flow collaborative warning strategy with the target warning area as the warning boundary. That is to say, by locating the warning area through warning intervention analysis, realizing early perception, identifying the risk factors in multiple areas and performing risk trajectory prediction, evaluating the predicted risks, selecting appropriate warning strategies for warning intervention, forming an intelligent collaborative warning system, improving the accuracy of pedestrian behavior prediction, and thus enhancing the pedestrian safety level in complex traffic environments.

[0017] The above description is only an overview of the technical solution of the present application. In order to better understand the technical means of the present application, it can be implemented according to the content of the specification. In order to make the above and other purposes, features, and advantages of the present application more obvious and understandable, the specific embodiments of the present application are specifically described below. It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present application, nor is it used to limit the scope of the present application. Other features of the present application will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings described below are only exemplary, and for those of ordinary skill in the art, other drawings can be obtained according to the provided drawings without creative efforts.

[0019] Figure 1 It is a schematic flow chart of a pedestrian safety warning method combining high-precision sensing technology in the present application;

[0020] Figure 2 It is a schematic structural diagram of a pedestrian safety warning device combining high-precision sensing technology in the present application.

[0021] Description of the reference numerals: warning intervention analysis module 11, environmental intermittent sensing module 12, risk factor positioning module 13, risk trajectory prediction module 14, risk level determination module 15, warning strategy matching module 16, collaborative warning module 17. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0022] By providing a pedestrian safety warning method and device combining high-precision sensing technology, the present application solves the technical problem in the prior art that due to sensing limitations, it is difficult to accurately identify sudden behaviors of pedestrians, resulting in poor warning effects and thus affecting the overall traffic safety. By positioning the warning area through warning intervention analysis, early perception is realized, risk factors in multiple areas are identified and risk trajectories are predicted, the predicted risks are evaluated, appropriate warning strategies are selected for warning intervention, and an intelligent collaborative warning system is formed, improving the accuracy of pedestrian behavior prediction and thus enhancing the pedestrian safety level in complex traffic environments.

[0023] Next, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. It should be understood that the present application is not limited by the example embodiments described herein. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts belong to the scope of protection of the present application. Additionally, it should be noted that for the sake of description, only the parts related to the present application rather than all are shown in the accompanying drawings.

[0024] Embodiment 1. Please refer to the attached Figure 1 , the present application provides a pedestrian safety warning method combined with high-precision sensing technology. Among them, the pedestrian safety warning method combined with high-precision sensing technology is executed by a pedestrian safety warning device combined with high-precision sensing technology. The pedestrian safety warning method combined with high-precision sensing technology specifically includes the following steps:

[0025] S100: Through warning intervention analysis of the target risk area, locate the target warning area.

[0026] Further, S100 of the present application includes:

[0027] Based on the target risk area, call network data to obtain driving speed characteristics; interact to obtain the warning intervention response duration; calculate the warning intervention response distance according to the warning intervention response duration and acceleration speed characteristics; take the target risk area as the starting point and the warning intervention response distance as the road network framing scale to perform an extended road network framing to obtain the target warning area.

[0028] Specifically, warning intervention analysis of the target risk area means pre-evaluating the area that requires pedestrian safety risk warning to determine the key warning area, that is, the area where safety intervention measures need to be taken. By calling network data for the target risk area, obtain the driving speed characteristics of this area from the vehicle networking system, that is, the speed distribution of vehicles in the target risk area, such as average vehicle speed, maximum vehicle speed, acceleration, braking deceleration, etc. Network data call refers to calling real-time data through vehicle networking (V2X), cloud computing or intelligent transportation system (ITS), including vehicle speed, pedestrian flow, road congestion status, etc.

[0029] Interact with the traffic management center or other relevant systems to obtain the warning intervention response duration, that is, the time from issuing a warning signal to the driver / autonomous driving system taking a response, including the driver's reaction time (usually 0.5 to 1.5 seconds), system calculation time, and the time for the vehicle to perform braking or avoidance operations. According to the warning intervention response duration and acceleration speed characteristics, calculate the minimum distance required from the vehicle receiving the warning signal to completing braking or avoidance. Simply put, the response distance = speed × response duration. The determination of the actual response distance may be more complex, involving braking deceleration, the current speed of the vehicle, etc., and it shall be subject to the actual situation. For example, if the vehicle is traveling at a speed of 50 km / h, after the driver fully reacts and brakes, it requires at least 38 meters to stop safely.

[0030] Take the determined target risk area as the starting point and the calculated warning intervention response distance as the radius, and perform an extended selection in the road grid to form a complete target warning area. The extension road network selection means taking the target risk area as the center and extending and selecting the surrounding road grids within a certain distance range (warning intervention response distance) to form a larger target warning area for comprehensively perceiving moving objects such as pedestrians and vehicles. For example, at the school gate during the peak period, all roads within a radius of 2 km may be included in the warning area to ensure safety. By accurately calculating the warning intervention response distance, the area where warnings need to be implemented can be determined more accurately, reducing false alarms and missed alarms.

[0031] S200: Preset a warning intervention window and perform intermittent environmental perception of the target warning area based on the warning intervention window to obtain multiple regional dynamic information sets.

[0032] Furthermore, S200 of the present application includes:

[0033] Call network data based on the target warning area to obtain the traffic flow characteristics of the extended road network and the pedestrian flow characteristics of the extended road network; compare the traffic flow characteristics of the extended road network with the perception function number table to obtain the first perception configuration strategy; compare the pedestrian flow characteristics of the extended road network with the perception function number table to obtain the second perception configuration strategy; output the target perception configuration strategy by fusing the first perception configuration strategy and the second perception configuration strategy; configure a high-precision perception array for the target warning area with the target perception configuration strategy as a constraint, where the high-precision perception array includes multiple perception nodes configured in multiple perception areas.

[0034] Specifically, for the target warning area determined through the previous-step analysis, the pedestrian and vehicle flows need to be monitored with emphasis. Similarly, network data is called, and real-time data, including vehicle flow characteristics and pedestrian flow characteristics, is obtained through V2X (vehicle-road cooperation), traffic management cloud platform, cameras, radars, etc. The vehicle flow characteristics of the extended road network are the characteristics of vehicle flow on the roads within and around the target warning area, such as flow rate, speed, vehicle flow density, etc. The pedestrian flow characteristics of the extended road network are the characteristics of pedestrian flow on the roads within and around the target warning area, such as the number of pedestrians, flow direction, speed, etc.

[0035] Compare the vehicle flow characteristics of the extended road network with the perception function table to obtain the first perception configuration strategy. For example, according to the vehicle flow characteristics, select and configure millimeter-wave radar sensors that can cover a large area. The perception function table is a pre-defined matching rule between perception devices and different traffic scenarios, used to automatically select the best perception device configuration. Similarly, compare the pedestrian flow characteristics of the extended road network with the perception function table to obtain the second perception configuration strategy. For example, since the pedestrian flow in this area is large and the staying time is long, panoramic cameras and infrared sensors are selected.

[0036] The first perception configuration strategy is a sensor configuration plan suitable for vehicle flow monitoring, and the second perception configuration strategy is a sensor configuration plan suitable for pedestrian monitoring. Combining the characteristics of vehicle flow and pedestrian flow, fuse the first perception configuration strategy and the second perception configuration strategy, and output the target perception configuration strategy to ensure the monitoring accuracy. Use the target perception configuration strategy as a constraint, and arrange perception devices at multiple perception nodes in different sub-areas of the target warning area to achieve three-dimensional monitoring and form a high-precision perception array. By comprehensively considering the vehicle flow and pedestrian flow characteristics and selecting appropriate perception devices, the accuracy of environmental monitoring is improved.

[0037] Furthermore, this application also includes the following steps:

[0038] Through the fusion analysis of the vehicle flow characteristics of the extended road network and the pedestrian flow characteristics of the extended road network, obtain the effective duration of warning intervention; localize the warning intervention window based on the effective duration of warning intervention; use the warning intervention window as a constraint to perform synchronous intermittent operation of multiple perception nodes in the high-precision perception array to obtain multiple regional dynamic perception sequences of the multiple perception areas; after performing temporal fusion on the multiple regional dynamic perception sequences, perform risk feature detection based on the fusion result to obtain the multiple regional dynamic information sets.

[0039] Specifically, a fusion analysis is carried out on the traffic flow characteristics and pedestrian flow characteristics of the extended road network to determine the effective duration of warning intervention, that is, to calculate the time required for a vehicle to decelerate from the current speed to a safe stop, so as to ensure timely warning. The effective duration of warning intervention refers to the time interval from detecting potential traffic risks to still being able to effectively avoid accidents after taking intervention measures, which depends on factors such as traffic flow characteristics, pedestrian flow characteristics, and signal control strategies. Due to the complex road network situation, it is necessary to dynamically adjust the warning window according to the real-time situation. Therefore, the effective duration of warning intervention is used for adjustment to make it conform to the actual road conditions, ensuring that the warning is neither too early (resulting in false alarms) nor too late (resulting in safety accidents).

[0040] Taking the localized warning intervention window as a constraint, multiple sensing nodes in the high-precision sensing array are synchronously and intermittently operated. The multiple sensing nodes are synchronously started and stopped at specific time intervals to reduce power consumption and optimize data collection. For example, a certain camera and radar can alternately collect data for 0.5 seconds per second to avoid unnecessary repeated calculations. All sensing nodes cooperate to form multiple regional dynamic sensing sequences, that is, a set of sensing data that changes over time.

[0041] Perform temporal fusion on multiple regional dynamic sensing sequences, align them both in time and space, and merge multiple groups of synchronized image sequences of multiple regions into a stitched sensing image of the entire region according to the mapping relationship with the target warning region. Detect traffic flow characteristics and pedestrian flow characteristics for the stitched sensing image of each region, and store the vehicle dynamic characteristics and pedestrian dynamic characteristics according to the sensing time of the stitched sensing image to obtain multiple regional dynamic information sets. For example, store the detected risk characteristics and the corresponding vehicle and pedestrian dynamic information in each region together to form a regional dynamic information set. Multiple regional dynamic information sets include the collection of information such as vehicles, pedestrians, weather, and light collected at different time points in different regions, which is used to analyze the dynamic changes within the region, such as the changing trends of traffic flow and pedestrian flow.

[0042] Generally speaking, a suitable time window is set in advance, usually generated based on historical data, to obtain a warning intervention window. According to the traffic flow characteristics and pedestrian flow characteristics of the extended road network called by the constructed target warning region, set the effective duration of warning intervention, and localize the warning intervention window to make it more in line with the current road conditions. Perform environmental intermittent sensing on the target warning region, set the sensing frequency, and monitor according to the frequency to reduce the amount of calculation and energy consumption, while avoiding redundant data. Finally, through the above specific process, multiple regional dynamic information sets are obtained to comprehensively analyze the risk status of the entire warning region and improve the intelligence level of safety warning.

[0043] Furthermore, this application also includes the following steps:

[0044] Perform spatio-temporal synchronization of the multiple regional dynamic perception sequences according to the warning intervention window to obtain multiple sets of regional dynamic perception sequences; perform multi-view image sequence synchronization stitching on the multiple sets of regional dynamic perception sequences according to the mapping relationship between the multiple perception regions and the target warning region, and output multiple regional stitched perception images; perform vehicle flow feature detection on the first regional stitched perception image to obtain K vehicle dynamic features of K real-time vehicles, where the vehicle dynamic features include vehicle identification features, driving speed features, and driving direction features; perform pedestrian flow feature detection on the first regional stitched perception image to obtain F pedestrian dynamic features of F real-time pedestrians, where the pedestrian dynamic features include pedestrian identification features, walking speed features, and walking direction features; use the first perception time of the first regional stitched perception image to perform associated storage of the K vehicle dynamic features and the F pedestrian dynamic features to obtain a first regional dynamic information set; and so on, and obtain the multiple regional dynamic information sets by performing risk feature detection on the multiple regional stitched perception images.

[0045] Specifically, according to the localized warning intervention window, perform spatio-temporal synchronization processing on multiple regional dynamic perception sequences, and uniformly align the perception data of multiple regions in the time axis and spatial range to ensure the accuracy and timeliness of analysis. Use timestamp alignment to ensure that all data is stored according to the same time scale. When different perception devices are out of sync, use a clock synchronization algorithm to ensure that all sensor times are consistent. Use a spatial mapping method to match the data of different perception devices according to geographical coordinates. For example, if A, B, and C are at different positions on the same road section, they are associated according to the road topology structure. After spatio-temporal synchronization processing, multiple sets of regional dynamic perception sequences are obtained, where the data of different perception nodes are aligned in time and space.

[0046] According to the mapping relationship between multiple perception regions and the target warning region, merge the image sequences from different perspectives into a coherent image sequence to provide more comprehensive visual information. For example, use image processing techniques, such as image stitching algorithms, to merge the image sequences from different perspectives into a coherent image sequence. Adopt feature extraction algorithms such as SIFT, ORB, and SURF to find the key points in the images, use the fast nearest neighbor matching algorithm to calculate the matching points of different visual images, and perform image stitching. After image stitching, multiple regional stitched perception images are obtained.

[0047] Arbitrarily select one of the spliced perception images from multiple regions as the first-region spliced perception image for traffic flow feature monitoring. That is, based on the spliced perception image of this region, the features of the traffic flow are detected in real time, and specific vehicles are identified, including vehicle identification features (such as license plate, vehicle type, speed, etc.), driving speed (average speed), and driving direction. For example, at an intersection, 3 vehicles are detected: Vehicle 1 is an SUV with a speed of 45 km / h and a direction from east to west; Vehicle 2 is a sedan with a speed of 60 km / h and a direction from south to north; Vehicle 3 is a bus with a speed of 35 km / h and a direction from north to south. According to the specific road conditions, the direction can be specified to the street name.

[0048] Based on the first-region spliced perception image, pedestrian flow feature detection is performed, and passing pedestrians are detected in real time, including pedestrian identification features (such as pedestrian gender, clothing color, characteristic features, etc.), walking speed features (average speed), and walking direction features.

[0049] Associate and store the K vehicle dynamic features of K real-time vehicles and the F pedestrian dynamic features of F real-time pedestrians. Association is performed according to the first perception time of the first-region spliced perception image. For example, the traffic flow and pedestrian flow data at twelve o'clock are in one group, and then the situation at the next moment is continued to be recorded and stored in the database. Associate the dynamic features of each vehicle and pedestrian with the specific timestamp when they appear in the image. By analogy, by repeating the above steps for the spliced perception images of multiple regions, multiple region dynamic information sets are obtained, including the vehicle dynamic features and pedestrian dynamic features of each region. Through multi-view image sequence synchronous splicing and multi-feature detection, the perception accuracy of traffic flow and pedestrian flow dynamics is improved.

[0050] S300: Perform temporal risk feature identification on the multiple region dynamic information sets based on the information perception time sequence to locate the distribution of dynamic risk factors.

[0051] Furthermore, S300 of this application includes:

[0052] Based on the consistency of vehicle recognition features, perform temporal aggregation of driving change features on the multiple regional dynamic information sets to obtain K first feature change sequences of the K real-time vehicles; based on the consistency of pedestrian recognition features, perform temporal aggregation of walking change features on the multiple regional dynamic information sets to obtain F second feature change sequences of the F real-time pedestrians; preset driving feature change scales, where the driving feature change scales include driving speed change scales and driving direction change scales; traverse the K first feature change sequences using the driving feature change scales to perform risk judgment on the K real-time vehicles and output M risk vehicles; by analogy, preset walking feature change scales to perform risk judgment on the F real-time pedestrians and output N risk pedestrians; extract M vehicle position features of the M risk vehicles and N pedestrian position features of the N risk pedestrians as the dynamic risk factor distribution.

[0053] Specifically, aggregate the driving change features of the same vehicle in the multiple regional dynamic information sets according to time, and merge and analyze the vehicle driving feature (such as speed, direction) data at different time points to track the behavior changes of the vehicle, obtaining K first feature change sequences of the K real-time vehicles, including the driving speed change features and driving direction change features of each vehicle. The consistency of vehicle recognition features means that within different times and different perception regions, the same vehicle is identified and confirmed through information such as license plate number, vehicle color, vehicle type, and body features.

[0054] According to traffic safety regulations and actual situations, preset driving feature change scales, including driving speed change scales and driving direction change scales, for evaluating the degree of vehicle behavior changes. For example, set the risk threshold that the vehicle speed change acceleration exceeds 20 km / h or the direction change exceeds 25 degrees within 1 s. According to the driving feature change scales, traverse the K first feature change sequences of the K real-time vehicles to judge whether they meet the risk criteria. All vehicles that exceed the preset scales are regarded as risk vehicles, obtaining M risk vehicles, which may have dangerous driving behaviors.

[0055] Similar to vehicle driving, aggregate the walking change features of the same person in the multiple regional dynamic information sets according to time, merge and analyze the pedestrian travel feature data at different time periods to track the behavior changes of the pedestrian, obtaining F second feature change sequences of the F real-time pedestrians, including the travel speed change features and travel direction change features of each person. The consistency of pedestrian recognition features is similar to vehicle recognition, using methods such as clothing color, body shape, gait features, and AI face recognition to confirm the trajectories of the same pedestrian at different time points.

[0056] According to traffic safety regulations and actual situations, preset the change scales of pedestrian characteristics, including the change scale of walking speed and the change scale of walking direction, which are used to evaluate the degree of change in pedestrian behavior. For example, the walking speed suddenly increases by 0.8 m / s, and the direction may suddenly cross the road within 1 s. Traverse all pedestrian data to detect whether it meets the risk criteria. All pedestrians who do not meet the preset scales are regarded as risk pedestrians, and N risk pedestrians are obtained, who may have dangerous behaviors such as crossing the road.

[0057] Extract the M vehicle position characteristics of M risk vehicles and the N pedestrian position characteristics of N risk pedestrians to form a dynamic risk factor distribution, that is, where and when there are high-risk vehicles and pedestrians, which is used for risk area modeling. Through time series aggregation and the preset feature change scales, accurately identify the vehicle and pedestrian behaviors that may lead to traffic accidents, respond to potential dangerous situations in advance, and issue warnings in a timely manner.

[0058] S400: Using the target risk area as the risk boundary, perform risk trajectory prediction on the dynamic risk factor distribution to obtain the early warning intervention risk distribution.

[0059] Furthermore, S400 of this application includes:

[0060] Extract the M first feature change sequences of the M risk vehicles; extract the N second feature change sequences of the N risk pedestrians; use the target risk area as the risk boundary and the target early warning area as the early warning boundary, and perform driving prediction on the M risk vehicles according to the M first feature change sequences to obtain the vehicle intervention risk distribution; use the target risk area as the risk boundary and the target early warning area as the early warning boundary, and perform walking prediction on the N risk pedestrians according to the N second feature change sequences to obtain the pedestrian intervention risk distribution; perform cross-fusion of the vehicle intervention risk distribution and the pedestrian intervention risk distribution based on the risk position distribution, and output the early warning intervention risk distribution.

[0061] Specifically, randomly extract the M first feature change sequences of M risk vehicles from the dynamic risk factor distribution, that is, the data sequences of vehicle driving characteristics (such as speed and direction) changing over time. Use the target risk area as the risk boundary and the target early warning area as the early warning boundary. The risk boundary refers to the area range defining the existence of risk, and the early warning boundary refers to the area range for issuing warnings. Use machine learning algorithms, such as time series prediction models, to analyze the driving feature change sequences of each risk vehicle and predict its future driving behavior. For example, by analyzing the speed and direction changes of the vehicle, predict whether the vehicle may speed or suddenly turn.

[0062] Select a suitable time - series prediction model, such as the ARIMA model or the LSTM network, to analyze the change sequence of the driving characteristics of each risk vehicle. Use the data of the past 2 seconds of each vehicle to predict the future 2 - second movement trajectory. For example, use the ARIMA model to analyze the changes in vehicle speed and direction over time and predict the future speed and direction of the vehicle. According to the analysis results of the time - series prediction model, predict the future driving behavior of each risk vehicle. For example, predict the changes in vehicle speed and direction in the next few seconds. Based on the predicted vehicle driving behavior, generate a vehicle risk distribution. For example, if a vehicle's speed decreases from 20 m / s to 5 m / s and the direction angle deviates by 15 degrees in the past 5 seconds, it is predicted that it may change lanes or stop at the next moment. The vehicle intervention risk distribution refers to the risk distribution map obtained by performing a spatial analysis on the driving prediction results of all M risk vehicles.

[0063] Similarly, perform the above steps for risk pedestrians, which will not be elaborated here. Randomly extract N second - feature change sequences of N risk pedestrians from the dynamic risk - factor distribution, that is, the data sequences of pedestrian driving characteristics (such as speed, direction) changing over time. Take the target risk area as the risk boundary and the target warning area as the warning boundary. According to the N second - feature change sequences, perform walking predictions for N risk pedestrians to obtain the pedestrian intervention risk distribution. For example, if it is found that a pedestrian stays at an intersection for more than 10 seconds and shows random direction changes, it may be due to blocked vision or distracted behavior, and it is predicted that the pedestrian may enter the lane within the next 3 seconds. The pedestrian intervention risk distribution refers to the risk distribution map obtained by performing a spatial analysis on the walking prediction results of all N risk pedestrians.

[0064] According to the determined risk - position distribution, that is, the position where each risk is located, cross - fuse the risk distributions of vehicles and pedestrians. Consider the position relationship between risk vehicles and pedestrians and their impact on traffic flow, and comprehensively obtain the warning - intervention risk distribution of the target area. The warning - intervention risk distribution is the final risk distribution map of the entire target area, which can help the traffic management system or the intelligent driving system to perform active warning and intervention. By extracting the feature change sequences of risk vehicles and pedestrians, performing driving and walking predictions, and cross - fusing based on the risk - position distribution, it provides a more efficient and accurate risk assessment and warning strategy for the intelligent transportation system, effectively improving traffic safety.

[0065] S500: Evaluate the risk level according to the warning - intervention risk distribution and output the real - time risk level.

[0066] S600: Match the warning strategy according to the real - time risk level and output the traffic - flow collaborative warning strategy.

[0067] Specifically, the risk distribution of early warning intervention obtained by the above calculation, that is, the traffic risk distribution map of the target area, shows the risk situation of each area at the current moment. The risk level of early warning intervention risk distribution is evaluated, and the risk areas are classified and graded to quantify the risk level. Using risk assessment methods such as the analytic hierarchy process (AHP) or the fuzzy comprehensive evaluation method, the early warning intervention risk distribution is analyzed and the risks are divided into different levels. According to the risk distribution, the risks are divided into three levels: high, medium, and low. According to the results of the risk level evaluation, the real-time risk level of each area is output. For example, if the risk of an area is assessed as high risk, a high risk level is output.

[0068] According to the real-time risk level, select the appropriate warning strategy, such as adjusting the duration of traffic lights, issuing driver reminders, and deploying traffic police patrols. The traffic flow collaborative warning strategy is an optimization strategy that comprehensively considers factors such as vehicles, pedestrians, traffic lights, and road control, aiming to reduce traffic congestion and accident risks. According to the matching results of the warning strategy, the traffic flow collaborative warning strategy is output. For example, if you choose to issue an emergency warning, you may need to coordinate traffic lights and vehicle communication systems to ensure the timely communication of warning information. For example, the current vehicle flow at a certain intersection is 200 vehicles / min, and the current pedestrian flow is 1320 people / min. The calculated risk value is: R = 0.88 (high risk), and the matching warning strategy is: the on-board navigation prompts high-risk areas, please drive carefully, recommend drivers to pass at a low speed, and adjust the traffic lights. At the same time, increase traffic police dispatch and robot patrols, and voice broadcasts to remind pedestrians to pay attention to safety. Through the real-time traffic risk level, the optimal warning strategy is dynamically matched, and a traffic flow collaborative optimization plan is formed, which provides a more efficient and accurate warning strategy for the intelligent transportation system and effectively improves traffic safety.

[0069] S700: Taking the target warning area as the warning boundary, the traffic flow collaborative warning strategy is adopted to perform traffic flow collaborative warning optimization for pedestrian safety.

[0070] Specifically, the target warning area is used as the warning boundary, that is, the road area that needs to be monitored and intervened, such as school gates, crossroads, shopping mall entrances and exits, etc., where there is a large pedestrian flow and high traffic risks. Implement a traffic flow collaborative warning strategy in the target area, combine factors such as pedestrians, vehicles, signal lights, and traffic management systems for overall optimization, improve road safety and reduce traffic accidents. For example, when there is less traffic but dense pedestrians, extend the green light time on the sidewalk. Set up an electronic fence, and when a pedestrian runs a red light or enters a dangerous area, the system automatically triggers a voice reminder. Send an alert to the connected vehicle (V2X) for a high-density pedestrian area to remind the driver to slow down. By setting the target warning area and using the traffic flow collaborative warning strategy, the passage of pedestrians and vehicles is optimized, and road safety and traffic efficiency are improved.

[0071] In summary, the pedestrian safety warning method combining high-precision perception technology provided by this application has the following beneficial effects:

[0072] By performing warning intervention analysis on the target risk area to locate the target warning area; presetting a warning intervention window, and performing intermittent environmental perception of the target warning area based on the warning intervention window to obtain multiple regional dynamic information sets; performing temporal risk feature recognition of the multiple regional dynamic information sets based on the information perception time sequence to locate the distribution of dynamic risk factors; using the target risk area as the risk boundary to perform risk trajectory prediction on the distribution of the dynamic risk factors to obtain a warning intervention risk distribution; performing risk level evaluation according to the warning intervention risk distribution to output a real-time risk level; performing warning strategy matching according to the real-time risk level to output a traffic flow collaborative warning strategy; using the target warning area as the warning boundary and adopting the traffic flow collaborative warning strategy to perform traffic flow collaborative warning optimization on pedestrian safety. That is to say, by performing warning intervention analysis to locate the warning area, realizing early perception, identifying the risk factors in multiple areas and performing risk trajectory prediction, evaluating the predicted risks, selecting appropriate warning strategies for warning intervention, forming an intelligent collaborative warning system, improving the accuracy of pedestrian behavior prediction, and thus enhancing the pedestrian safety level in complex traffic environments.

[0073] Embodiment 2. Based on the same inventive concept as the pedestrian safety warning method combining high-precision perception technology in the foregoing Embodiment 1, this application also provides a pedestrian safety warning device combining high-precision perception technology. Please refer to the appendix Figure 2 The pedestrian safety warning device combining high-precision perception technology includes:

[0074] A warning intervention analysis module 11, configured to locate a target warning area by performing warning intervention analysis on a target risk area; an environmental intermittent perception module 12, configured to preset a warning intervention window and perform intermittent environmental perception of the target warning area based on the warning intervention window to obtain multiple regional dynamic information sets; a risk factor location module 13, configured to perform temporal risk feature recognition of the multiple regional dynamic information sets based on the information perception time sequence to locate the distribution of dynamic risk factors; a risk trajectory prediction module 14, configured to use the target risk area as the risk boundary to perform risk trajectory prediction on the distribution of the dynamic risk factors to obtain a warning intervention risk distribution; a risk level determination module 15, configured to perform risk level evaluation according to the warning intervention risk distribution to output a real-time risk level; a warning strategy matching module 16, configured to perform warning strategy matching according to the real-time risk level to output a traffic flow collaborative warning strategy; a collaborative warning module 17, configured to use the target warning area as the warning boundary and adopt the traffic flow collaborative warning strategy to perform traffic flow collaborative warning optimization on pedestrian safety.

[0075] Furthermore, the warning intervention analysis module 11 in the pedestrian safety warning device combined with high-precision sensing technology is further configured to:

[0076] Based on the target risk area, call network data to obtain driving speed characteristics; interact to obtain the warning intervention response duration; calculate the warning intervention response distance according to the warning intervention response duration and the acceleration speed characteristics; use the target risk area as the starting point and the warning intervention response distance as the road network selection scale for extension road network selection to obtain the target warning area.

[0077] Furthermore, the environmental intermittent sensing module 12 in the pedestrian safety warning device combined with high-precision sensing technology is further configured to:

[0078] Based on the target warning area, call network data to obtain the vehicle flow characteristics of the extended road network and the pedestrian flow characteristics of the extended road network; compare the vehicle flow characteristics of the extended road network with the sensing function number table to obtain the first sensing configuration strategy; compare the pedestrian flow characteristics of the extended road network with the sensing function number table to obtain the second sensing configuration strategy; output the target sensing configuration strategy by fusing the first sensing configuration strategy and the second sensing configuration strategy; configure the high-precision sensing array in the target warning area with the target sensing configuration strategy as a constraint, where the high-precision sensing array includes multiple sensing nodes configured in multiple sensing areas.

[0079] Furthermore, the environmental intermittent sensing module 12 in the pedestrian safety warning device combined with high-precision sensing technology is further configured to:

[0080] Obtain the effective warning intervention duration by fusing and analyzing the vehicle flow characteristics of the extended road network and the pedestrian flow characteristics of the extended road network; localize the warning intervention window based on the effective warning intervention duration; synchronously and intermittently operate the multiple sensing nodes in the high-precision sensing array with the warning intervention window as a constraint to obtain multiple regional dynamic sensing sequences of the multiple sensing areas; after performing temporal fusion on the multiple regional dynamic sensing sequences, perform risk feature detection based on the fusion result to obtain the multiple regional dynamic information sets.

[0081] Furthermore, the environmental intermittent sensing module 12 in the pedestrian safety warning device combined with high-precision sensing technology is further configured to:

[0082] Synchronize the spatio-temporal of the multiple regional dynamic perception sequences according to the warning intervention window to obtain multiple groups of regional dynamic perception sequences; perform multi-view image sequence synchronization stitching on the multiple groups of regional dynamic perception sequences according to the mapping relationship between the multiple perception regions and the target warning region, and output multiple regional stitched perception images; perform vehicle flow feature detection on the first regional stitched perception image to obtain K vehicle dynamic features of K real-time vehicles, where the vehicle dynamic features include vehicle identification features, driving speed features, and driving direction features; perform pedestrian flow feature detection on the first regional stitched perception image to obtain F pedestrian dynamic features of F real-time pedestrians, where the pedestrian dynamic features include pedestrian identification features, walking speed features, and walking direction features; use the first perception time of the first regional stitched perception image to perform associated storage of the K vehicle dynamic features and the F pedestrian dynamic features to obtain a first regional dynamic information set; and so on, perform risk feature detection on the multiple regional stitched perception images to obtain the multiple regional dynamic information sets.

[0083] Further, the risk factor positioning module 13 in the pedestrian safety warning device combined with high-precision perception technology is further configured to:

[0084] Based on the consistency of vehicle identification features, perform temporal aggregation of driving change features on the multiple regional dynamic information sets to obtain K first feature change sequences of the K real-time vehicles; based on the consistency of pedestrian identification features, perform temporal aggregation of walking change features on the multiple regional dynamic information sets to obtain F second feature change sequences of the F real-time pedestrians; preset a driving feature change scale, where the driving feature change scale includes a driving speed change scale and a driving direction change scale; use the driving feature change scale to traverse the K first feature change sequences to perform risk judgment on the K real-time vehicles, and output M risk vehicles; and so on, preset a walking feature change scale to perform risk judgment on the F real-time pedestrians, and output N risk pedestrians; extract M vehicle position features of the M risk vehicles and N pedestrian position features of the N risk pedestrians as the dynamic risk factor distribution.

[0085] Further, the risk trajectory prediction module 14 in the pedestrian safety warning device combined with high-precision perception technology is further configured to:

[0086] Extract the M first feature change sequences of the M risk vehicles; extract the N second feature change sequences of the N risk pedestrians; use the target risk area as the risk boundary and the target warning area as the warning boundary, and perform driving predictions on the M risk vehicles according to the M first feature change sequences to obtain the vehicle intervention risk distribution; use the target risk area as the risk boundary and the target warning area as the warning boundary, and perform walking predictions on the N risk pedestrians according to the N second feature change sequences to obtain the pedestrian intervention risk distribution; perform cross-fusion of the vehicle intervention risk distribution and the pedestrian intervention risk distribution based on the risk position distribution, and output the warning intervention risk distribution.

[0087] The various embodiments in this specification are described in a progressive manner. The key point of each embodiment is to illustrate the differences from other embodiments. The Figure 1 The pedestrian safety warning method and specific examples combined with high-precision sensing technology in the first embodiment are equally applicable to the pedestrian safety warning device combined with high-precision sensing technology in this embodiment. Through the detailed description of the pedestrian safety warning method combined with high-precision sensing technology above, those skilled in the art can clearly understand the pedestrian safety warning device combined with high-precision sensing technology in this embodiment. Therefore, for the sake of simplicity of the specification, it will not be elaborated here. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple. For the relevant parts, please refer to the description in the method section.

[0088] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0089] Obviously, those skilled in the art can make several improvements and modifications to the present application without departing from the principles of the present application, and these improvements and modifications also fall within the protection scope of the present application.

Claims

1. A pedestrian safety warning method combined with high-precision sensing technology, characterized in that: include: Locate the target warning area by conducting early warning intervention analysis on the target risk area; Presetting an early warning intervention window, and performing intermittent environmental perception of the target early warning area based on the early warning intervention window to obtain multiple regional dynamic information sets; Based on the information perception time series, the time series risk characteristics of the multiple regional dynamic information sets are identified to locate the distribution of dynamic risk factors; Taking the target risk area as the risk boundary, predicting the risk trajectory of the dynamic risk factor distribution to obtain the early warning intervention risk distribution; Perform risk level evaluation based on the early warning intervention risk distribution and output real-time risk level; Matching early warning strategies according to the real-time risk level and outputting traffic flow collaborative early warning strategies; The target warning area is used as the warning boundary, and the traffic flow collaborative warning strategy is adopted to perform traffic flow collaborative warning optimization for pedestrian safety.

2. The pedestrian safety warning method combined with high-precision sensing technology as claimed in claim 1, characterized in that: By conducting early warning intervention analysis on target risk areas, the target early warning areas are located, including: Based on the target risk area, network data is called to obtain driving speed characteristics; Interactively obtain the warning intervention response time; Calculating and obtaining the early warning intervention response distance according to the early warning intervention response duration and acceleration speed characteristics; Taking the target risk area as the starting point and the early warning intervention response distance as the road network selection scale, an extended road network selection is performed to obtain the target early warning area.

3. The pedestrian safety warning method combined with high-precision sensing technology as claimed in claim 2, characterized in that: Based on the early warning intervention window, intermittent environmental perception of the target early warning area is performed to obtain multiple regional dynamic information sets, and before that, it also includes: Based on the target warning area, network data is called to obtain the vehicle flow characteristics and pedestrian flow characteristics of the extended road network; The vehicle flow characteristics of the extended road network are compared with the perception function table to obtain a first perception configuration strategy; The second perception configuration strategy is obtained by comparing the pedestrian flow characteristics of the extended road network with the perception function table; Outputting a target perception configuration strategy by fusing the first perception configuration strategy and the second perception configuration strategy; The high-precision sensing array of the target warning area is configured with the target sensing configuration strategy as a constraint, wherein the high-precision sensing array includes a plurality of sensing nodes configured in a plurality of sensing areas.

4. The pedestrian safety warning method combined with high-precision sensing technology as claimed in claim 3, characterized in that: Based on the early warning intervention window, intermittent environmental perception of the target early warning area is performed to obtain multiple regional dynamic information sets, further comprising: By integrating and analyzing the vehicle flow characteristics and the pedestrian flow characteristics of the extended road network, the effective duration of early warning intervention is obtained; Localizing the early warning intervention window based on the effective duration of the early warning intervention; Taking the early warning intervention window as a constraint, performing synchronous intermittent operation of multiple sensing nodes in the high-precision sensing array to obtain multiple regional dynamic sensing sequences of the multiple sensing areas; After performing time series fusion on the multiple regional dynamic perception sequences, risk feature detection is performed based on the fusion results to obtain the multiple regional dynamic information sets.

5. The pedestrian safety warning method combined with high-precision sensing technology as claimed in claim 4, characterized in that: After performing time series fusion on the multiple regional dynamic perception sequences, risk feature detection is performed based on the fusion result to obtain the multiple regional dynamic information sets, further comprising: Performing spatiotemporal synchronization of the multiple regional dynamic perception sequences according to the early warning intervention window to obtain multiple groups of regional dynamic perception sequences; Synchronously stitching the multi-view image sequences of the multiple groups of regional dynamic perception sequences according to the mapping relationship between the multiple perception areas and the target warning areas, and outputting multiple regional stitching perception images; Performing traffic flow feature detection on the first region stitched perception image to obtain K vehicle dynamic features of K real-time vehicles, wherein the vehicle dynamic features include vehicle identification features, driving speed features, and driving direction features; Performing crowd flow feature detection on the first region stitched perception image to obtain F pedestrian dynamic features of F real-time pedestrians, wherein the pedestrian dynamic features include pedestrian recognition features, walking speed features, and walking direction features; Using the first perception time of the first region stitching perception image to associate and store K vehicle dynamic features and F pedestrian dynamic features, to obtain a first region dynamic information set; By analogy, the multiple regional dynamic information sets are obtained by performing risk feature detection on the multiple regional spliced ​​perception images.

6. The pedestrian safety warning method combined with high-precision sensing technology as claimed in claim 5, characterized in that: Based on the information perception time series, the time series risk characteristics of the multiple regional dynamic information sets are identified to locate the dynamic risk factor distribution, and further include: Based on the consistency of vehicle identification features, performing driving change feature time series aggregation on the multiple regional dynamic information sets to obtain K first feature change sequences of the K real-time vehicles; Based on the consistency of pedestrian recognition features, performing walking change feature time series aggregation on the multiple regional dynamic information sets to obtain F second feature change sequences of the F real-time pedestrians; Presetting a driving characteristic change scale, wherein the driving characteristic change scale includes a driving speed change scale and a driving direction change scale; Using the driving characteristic change scale to traverse the K first characteristic change sequences, perform risk judgment on the K real-time vehicles, and output M risky vehicles; Similarly, the walking characteristic change scale is preset to perform risk judgment on the F real-time pedestrians, and N risky pedestrians are output; The M vehicle position features of the M risky vehicles and the N pedestrian position features of the N risky pedestrians are extracted as the dynamic risk factor distribution.

7. The pedestrian safety warning method combined with high-precision sensing technology as claimed in claim 6, characterized in that: Taking the target risk area as the risk boundary, predicting the risk trajectory of the dynamic risk factor distribution to obtain the early warning intervention risk distribution also includes: Extracting M first feature change sequences of the M risky vehicles; Extracting N second feature change sequences of the N risky pedestrians; Taking the target risk area as a risk boundary and the target warning area as a warning boundary, performing driving prediction of the M risky vehicles according to the M first feature change sequences to obtain a vehicle intervention risk distribution; Taking the target risk area as a risk boundary and the target warning area as a warning boundary, performing walking prediction of the N risky pedestrians according to the N second feature change sequences to obtain a pedestrian intervention risk distribution; The vehicle intervention risk distribution and the pedestrian intervention risk distribution are cross-fused based on the risk location distribution, and the early warning intervention risk distribution is output.

8. A pedestrian safety warning device combined with high-precision sensing technology, characterized in that: The steps for implementing the pedestrian safety warning method combined with high-precision sensing technology as described in any one of claims 1 to 7, wherein the pedestrian safety warning device combined with high-precision sensing technology comprises: The early warning intervention analysis module is used to locate the target early warning area by conducting early warning intervention analysis on the target risk area; An environment intermittent perception module, used to preset an early warning intervention window, and perform environment intermittent perception of the target early warning area based on the early warning intervention window to obtain multiple regional dynamic information sets; A risk factor positioning module, used to identify the time series risk characteristics of the multiple regional dynamic information sets based on the information perception time series, and locate the dynamic risk factor distribution; A risk trajectory prediction module is used to predict the risk trajectory of the dynamic risk factor distribution with the target risk area as the risk boundary to obtain the early warning intervention risk distribution; A risk level determination module, used to evaluate the risk level according to the early warning intervention risk distribution and output a real-time risk level; A warning strategy matching module, used to match warning strategies according to the real-time risk level and output a traffic flow collaborative warning strategy; The collaborative warning module is used to use the target warning area as the warning boundary and adopt the traffic flow collaborative warning strategy to perform traffic flow collaborative warning optimization for pedestrian safety.

Images