Pedestrian collision risk microscopic analysis method and system based on multivariate Bayesian space model

Abstract

The invention relates to the technical field of pedestrian traffic safety, and discloses a pedestrian collision risk microscopic analysis method and system based on a multivariate Bayesian space model.The method comprises the steps that road pedestrian collision accident data are collected, accidents are divided into light injury, heavy injury and death, the accident data are screened, and accident feature data are extracted; the number of people is detected in real time through monitoring data, and real-time pedestrian data in two directions of a road are obtained; road static information and road dynamic information are obtained, and road feature data and traffic feature data of a road are extracted respectively; and constructing a multivariate Bayesian space model, and evaluating the road pedestrian collision risk based on the pedestrian data, the road feature data and the traffic feature data. According to the method, the accident feature data, the road feature data and the traffic feature data are extracted, the real-time pedestrian data are combined, the multivariate Bayesian space model is used for analysis, and the road pedestrian collision risk can be comprehensively and meticulously evaluated.

Description

Technical Field

[0001] The present invention relates to the technical field of pedestrian traffic safety, and in particular to a microscopic analysis method and system for pedestrian collision risk based on a multivariate Bayesian space model. Background Art

[0002] The study of traffic accident influencing factors is essentially a regression problem. At the macro level, there is a significant correlation between the change rates of influencing factors such as gross domestic product, automobile ownership, population density, traffic investment, and road mileage and the growth rate of road traffic accident fatalities. However, at the macro level, only the overall laws and characteristics of accident data are analyzed, or the causal risks of traffic accidents are studied under specific conditions. There is less research on the difference analysis of significant influencing factors between accidents of different severity levels, resulting in the weakening or hiding of its significant influencing factors.

[0003] At the micro level, potential influencing factors can be explored, including road characteristics, traffic characteristics, etc., so as to more reasonably evaluate and prevent traffic safety risks. Street space plays a crucial role in pedestrian safety, but the impact of fine-scale street environmental characteristics has not been fully studied. The present invention studies the impact of various factors on pedestrian traffic accidents from the micro scale of the street level, so as to provide a method for urban planning to reduce pedestrian collision risks and promote a safer street environment conducive to walking. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed. Therefore, the present invention provides a microscopic analysis method for pedestrian collision risk based on a multivariate Bayesian space model to solve the above problems.

[0005] To solve the above technical problems, the present invention provides the following technical solutions: In a first aspect, the present invention provides a microscopic analysis method for pedestrian collision risk based on a multivariate Bayesian space model, including: Collect road pedestrian collision accident data. The accidents are divided into three types: minor injury, serious injury, and death. Screen the accident data and extract accident feature data; Real-time detect the number of pedestrians through monitoring data and obtain real-time pedestrian data in two directions of the road; Obtain road static information and road dynamic information, and respectively extract road feature data and traffic feature data of the road; Construct a multivariate Bayesian space model, and evaluate the road pedestrian collision risk based on the pedestrian data, road feature data, and traffic feature data.

[0006] As a preferred scheme of the microscopic analysis method for pedestrian collision risk based on the multivariate Bayesian space model of the present invention, wherein: extracting accident feature data includes: Obtain road pedestrian collision accident data, screen the accident data, eliminate non-motor vehicle and pedestrian collision accidents, and only analyze motor vehicle and pedestrian collision accidents; Based on the road pedestrian collision accident data, extract information associated with traffic accidents as features, including accident occurrence time, accident occurrence location, age, gender, and weather conditions.

[0007] As a preferred embodiment of the microscopic analysis method for pedestrian collision risk based on a multivariate Bayesian spatial model according to the present invention, wherein: obtaining real-time pedestrian data in two directions of the road includes: Through real-time video streams, use the YOLOv8 algorithm for object recognition to obtain pedestrian count data in two directions of the road; Through time window division, discretize the data into hourly data and daily data in segments, and then perform window aggregation to count the number of pedestrians in each direction.

[0008] As a preferred embodiment of the microscopic analysis method for pedestrian collision risk based on a multivariate Bayesian spatial model according to the present invention, wherein: extracting road feature data of the road includes: Through real-time video streams, obtain road static information and extract road feature data of the accident-occurring road; The road feature data includes road length, road width, road condition, road speed limit, bus stop, and crosswalk, where the road condition refers to the smoothness of the road surface.

[0009] As a preferred embodiment of the microscopic analysis method for pedestrian collision risk based on a multivariate Bayesian spatial model according to the present invention, wherein: extracting traffic feature data of the road includes: Output bounding boxes through the YOLOv8 vehicle detection model of video frames in video data, assign unique IDs in combination with a tracking algorithm, count the number of non-repeating IDs within a set time window, and generate traffic flow data; Calculate the actual displacement based on the pixel displacement of the vehicle in consecutive video frames and the calibrated scale, and divide it by the frame interval time to obtain the vehicle speed; Based on the speed data of all vehicles, obtain the maximum vehicle speed and the 85th percentile vehicle speed.

[0010] As a preferred embodiment of the microscopic analysis method for pedestrian collision risk based on a multivariate Bayesian spatial model according to the present invention, wherein: constructing a multivariate Bayesian spatial model includes: Construct a Poisson lognormal regression to establish a quantitative relationship between accident severity and road covariates. At the same time, through spatial random effects and non-spatial random effects, separate unobserved spatial heterogeneity and non-spatial heterogeneity; Through a multi-variable model, the correlation between the counts of different injury severities is analyzed, and the random parameters are expressed as: ; wherein, represents the multi-variable fixed parameters, represents the normally distributed random term.

[0011] As a preferred embodiment of the microscopic pedestrian collision risk analysis method based on the multi-variable Bayesian space model of the present invention, wherein: the Poisson log-normal regression is expressed as: ; wherein, is the number of pedestrian casualties with the accident severity level of observed on the road section , representing minor injury, serious injury and death respectively when , ; ; wherein, represents the vector of covariates, is the intercept, is the parameter vector, represents the over-dispersed unstructured random term, represents the structured random term of spatial dependence, represents the total random effect term of the road section i under the accident severity k .

[0012] In a second aspect, the present invention provides a microscopic pedestrian collision risk analysis system based on a multi-variable Bayesian space model, comprising: An accident data preprocessing module, configured to collect road pedestrian collision accident data, the accidents are divided into three types: minor injury, serious injury and death, screen the accident data, and extract accident feature data; A pedestrian data module, configured to detect the number of pedestrians in real time through monitoring data and obtain real-time pedestrian data in two directions of the road; A road traffic feature acquisition module, configured to acquire road static information and road dynamic information, and respectively extract road feature data and traffic feature data of the road; A pedestrian collision risk analysis module, configured to construct a multi-variable Bayesian space model, and evaluate the road pedestrian collision risk based on the pedestrian data, road feature data and traffic feature data.

[0013] In a third aspect, the present invention provides a computer device, comprising: A memory and a processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the microscopic analysis method for pedestrian collision risk based on the multivariate Bayesian space model are implemented.

[0014] In a fourth aspect, the present invention provides a computer-readable storage medium storing computer-executable instructions, and when the computer-executable instructions are executed by a processor, the steps of the microscopic analysis method for pedestrian collision risk based on the multivariate Bayesian space model are implemented.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: By collecting accident data and classifying it, the present invention can accurately distinguish pedestrian collision accidents of different severity levels, which helps to deeply understand the characteristics of accidents. At the same time, by extracting accident feature data, road feature data, and traffic feature data, and combining real-time pedestrian data, and analyzing using the multivariate Bayesian space model, the road pedestrian collision risk can be comprehensively and meticulously evaluated, taking into account the comprehensive influence of various factors on the risk, which is more accurate than single-factor analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following-described drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0017] Figure 1 It is a schematic diagram of the overall process of the microscopic analysis method for pedestrian collision risk based on the multivariate Bayesian space model according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] In order to make the above objects, features, and advantages of the present invention more obvious and understandable, the following will make a detailed description of the specific embodiments of the present invention with reference to the drawings of the specification. Obviously, the described embodiments are some embodiments of the present invention, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0019] Referring to Figure 1 , an embodiment of the present invention provides a microscopic analysis method for pedestrian collision risk based on a multivariate Bayesian space model, including: S101, collecting road pedestrian collision accident data. The accidents are divided into three types: minor injury, serious injury, and death. The accident data is screened to extract accident feature data; S102, Detect the pedestrian flow in real time through monitoring data, and obtain the real-time pedestrian data in both directions of the road; S103, Obtain the static information and dynamic information of the road, and extract the road feature data and traffic feature data of the road respectively; S104, Construct a multivariate Bayesian space model, and evaluate the pedestrian collision risk on the road based on the pedestrian data, road feature data and traffic feature data.

[0020] Preferably, the accident feature data extraction includes: Obtain the road pedestrian collision accident data, screen the accident data, eliminate the collision accidents between non-motor vehicles and pedestrians, and only analyze the collision accidents between motor vehicles and pedestrians; Based on the road pedestrian collision accident data, extract the information associated with traffic accidents as features, including the accident occurrence time, accident occurrence location, age, gender, weather conditions.

[0021] Specifically, through the public security traffic police platform, collect the road pedestrian collision accident data around schools. According to the grading standards of the World Health Organization, the accidents are divided into three types: minor injuries, serious injuries and deaths. Minor injuries refer to those that require medical treatment but are not life-threatening and do not affect long-term living ability; serious injuries refer to those that may cause disabilities, organ function loss or endanger life; deaths refer to those that directly cause the death of personnel at the scene or within 24 hours after being sent to the hospital. Thus, calculate the probability distribution of accidents of different severity levels; screen the accident data, eliminate the collision accidents between non-motor vehicles and pedestrians, and only analyze the collision accidents between motor vehicles and pedestrians. Extract some information related to traffic accidents in the data as features, including the accident occurrence time, accident occurrence location, age, gender, weather conditions, etc.

[0022] Preferably, the real-time pedestrian data obtained in both directions of the road includes: Through the real-time video stream, use the YOLOv8 algorithm for target recognition to obtain the pedestrian counting data in both directions of the road; Through time window division, the data is discretized into hourly data and daily data in segments, and then window aggregation is performed to count the number of pedestrians in each direction.

[0023] It should be noted that by detecting the pedestrian flow in real time through monitoring data and obtaining the real-time pedestrian data, it is possible to timely grasp the dynamic situation of pedestrians on the road, evaluate the pedestrian collision risk in real time, and facilitate relevant departments to take timely measures, such as strengthening traffic guidance and setting warning signs during peak pedestrian flow periods, so as to effectively prevent accidents and reduce the pedestrian collision risk.

[0024] Preferably, the road feature data extraction of the road includes: Obtain road static information through real-time video streams, and extract road feature data of the road where the accident occurred; The road feature data includes road length, road width, road condition, road speed limit, bus stops, and crosswalks. Among them, the road condition refers to the smoothness of the road surface.

[0025] Specifically, obtain real-time video streams through video devices installed on each road and intersections by the public security traffic police platform, and obtain road feature data of the road where the accident occurred, including road length, road width, road condition, etc. Vehicle bumps caused by poor road conditions will lead to a decrease in the vehicle's evasive ability. Obtain boolean value data on whether there are bus stops and crosswalks on the road through POI data of the map open data platform.

[0026] Preferably, the extracted traffic feature data includes: Output bounding boxes through the YOLOv8 vehicle detection model of video frames in video data, assign unique IDs in combination with a tracking algorithm, count the number of non-repeating IDs within a set time window, and generate traffic flow data; Calculate the actual displacement based on the pixel displacement of the vehicle in consecutive video frames and the calibrated scale, and divide it by the frame interval time to obtain the vehicle speed; Based on the speed data of all vehicles, obtain the maximum vehicle speed and the 85th percentile vehicle speed.

[0027] Specifically, obtain video data through the intersection camera equipment of the public security traffic police platform, and obtain the road speed limit value through the map open platform.

[0028] Preferably, constructing a multivariate Bayesian spatial model includes: Construct a Poisson log-normal regression to establish a quantitative relationship between the accident severity and road covariates. At the same time, through spatial random effects and non-spatial random effects, separate the unobserved spatial heterogeneity and non-spatial heterogeneity; Through a multivariate model, analyze the correlation between the counts of different injury severities, and the random parameter is expressed as: ; Among them, represents the multivariate fixed parameter, represents the normal distribution random term.

[0029] Preferably, the Poisson log-normal regression is expressed as: ; Among them, is the number of pedestrian casualties with the accident severity level of observed on the road section ; represent minor injury, serious injury and death respectively when is the Poisson parameter, expressed as: ; ; where represents the vector of covariates, is the intercept, is the parameter vector, represents the unstructured random term of overdispersion, represents the structured random term of spatial dependence, represents the road section i under the accident severity k of the total random effect term.

[0030] To explain the influence of unobserved heterogeneity between observation units, the random parameter can be specified as: ; where is the fixed parameter, while is the random distribution error term subject to a normal distribution, with a mean of 0 and a variance of , expressed as: ; To explain the influence of spatial dependence, the conditional autoregression (CAR) given a priori will be adopted as follows: ; ; ; where represents the spatial weight, when, it means that road sections i and j are adjacent, otherwise 0; refers to the number of adjacent units adjacent to i, and the hyperparameter follows a Gamma distribution, represents the set of spatial random effects of all other road sections under the accident severity k excluding road section i, represents the conditional mean of road section i under the accident severity k, represents the spatial random effect of road section j under the accident severity k.

[0031] Through Poisson lognormal regression, establish the quantitative relationship between accident severity (minor injury / serious injury / death) and road covariates, capture spatial dependence, and at the same time, through the spatial random effect and non-spatial random effect , separate the unobserved spatial heterogeneity from the non-spatial heterogeneity.

[0032] It should be noted that the multivariate Bayesian spatial model takes into account spatial factors, can analyze the differences in pedestrian collision risks on roads at different locations, helps to identify accident-prone areas or high-risk regions, provides accurate basis for road planning and traffic facility improvement, and improves road safety.

[0033] Optionally, a multivariate model is used to consider the correlation between the counts of different injury severities m, and the random parameters can be specified as: ; where is the multivariate fixed parameter, is the normal distribution random term. In addition, the unstructured random term follows a multivariate normal distribution, and the unstructured random term can be expressed as: ; where is the mean vector, and is the covariance matrix estimated using the Wishart distribution, given by: ; where is the scale matrix of the precision matrix, is the degree of freedom. In this embodiment , the non-informative prior is set to: ; In addition, the multivariate CAR prior can be specified as: ; ; where represents the multivariate spatial random effect vector of section i, represents the set of spatial random effects of all other sections excluding section i, represents the covariance matrix of the multivariate spatial effects, , , respectively represent the spatial random effects of section j in the three dimensions of minor injury, serious injury, and death.

[0034] Estimate the covariance matrix through the multivariate normal distribution (MN) and Wishart prior, and reveal the latent interaction between variables, such as the combined effect of "road condition + speed" on mortality.

[0035] Estimate univariate and multivariate random parameter Poisson lognormal models using a Bayesian framework and Markov chain Monte Carlo simulation. The first 10,000 iterations are discarded as burn-in, and the next 20,000 iterations are run for each chain. Use 95% Bayesian credible intervals (BCIs) to evaluate the credibility of variables. To evaluate model fit, adopt the deviance information criterion (DIC), a generalized performance metric. The calculation method of DIC is as follows: ; where, is the posterior mean of the deviance, is the number of effective parameters. DIC takes into account both predictive performance ( ) and model complexity ( ).

[0036] It should be noted that by fitting the model, the credibility of variables within the 95% Bayesian credible interval (BCI) can be obtained, thereby exploring the correlation of each variable with pedestrian casualties. For example, through the model, it is identified that among the 3 high-risk sections, the road condition has the greatest impact on pedestrian casualties. Based on the results, corresponding traffic management measures can be formulated to control variables, such as re-paving the road surface, thereby reducing the risk of pedestrian collisions and promoting the construction of a safer street environment conducive to walking.

[0037] By collecting accident data and classifying them, the present invention can accurately distinguish pedestrian collision accidents of different severities, which helps to deeply understand the accident characteristics. At the same time, by extracting accident feature data, road feature data, and traffic feature data, and combining real-time pedestrian data, and using a multivariate Bayesian spatial model for analysis, the road pedestrian collision risk can be comprehensively and meticulously evaluated, taking into account the comprehensive impact of multiple factors on the risk, which is more accurate than single-factor analysis.

[0038] The above is a schematic solution of a microscopic analysis method for pedestrian collision risk based on a multivariate Bayesian spatial model in this embodiment. It should be noted that the technical solution of the microscopic analysis system for pedestrian collision risk based on the multivariate Bayesian spatial model belongs to the same concept as the above-mentioned microscopic analysis method for pedestrian collision risk based on the multivariate Bayesian spatial model. For the details not described in detail in the technical solution of the microscopic analysis system for pedestrian collision risk based on the multivariate Bayesian spatial model in this embodiment, reference can be made to the description of the technical solution of the microscopic analysis method for pedestrian collision risk based on the multivariate Bayesian spatial model.

[0039] This embodiment also provides a microscopic analysis system for pedestrian collision risk based on a multivariate Bayesian spatial model, including: An accident data preprocessing module, which is used to collect road pedestrian collision accident data. The accidents are divided into three types: minor injury, serious injury and death. It screens the accident data and extracts accident characteristic data; A pedestrian data module, which is used to detect the number of pedestrians in real time through monitoring data and obtain real-time pedestrian data in two directions of the road; A road traffic characteristics acquisition module, which is used to obtain road static information and road dynamic information, and extract road characteristic data and traffic characteristic data of the road respectively; A pedestrian collision risk analysis module, which is used to construct a multivariate Bayesian space model and evaluate the road pedestrian collision risk based on pedestrian data, road characteristic data and traffic characteristic data.

[0040] This embodiment also provides a computer device, which is applicable to the situation of microscopic analysis of pedestrian collision risk based on a multivariate Bayesian space model, including: A memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the method for microscopic analysis of pedestrian collision risk based on a multivariate Bayesian space model as proposed in the above embodiment.

[0041] This embodiment also provides a storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the method for microscopic analysis of pedestrian collision risk based on a multivariate Bayesian space model as proposed in the above embodiment.

[0042] The storage medium proposed in this embodiment and the method for microscopic analysis of pedestrian collision risk based on a multivariate Bayesian space model proposed in the above embodiment belong to the same inventive concept. For technical details not described in detail in this embodiment, reference can be made to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0043] From the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software and necessary general-purpose hardware. Of course, it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, in essence, or the part that makes a contribution to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as a floppy disk, a read-only memory (ROM), a random access memory (RAM), a flash memory (FLASH), a hard disk or an optical disc of a computer, etc., including several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of the various embodiments of the present invention.

[0044] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A pedestrian collision risk micro-analysis method based on a multivariate Bayesian spatial model, characterized in that: include: Collect road pedestrian collision accident data, which are divided into three types: minor injury, serious injury and death. Screen the accident data and extract accident feature data; Through monitoring data, the number of pedestrians can be detected in real time, and real-time pedestrian data in both directions of the road can be obtained; Obtaining road static information and road dynamic information, and extracting road characteristic data and traffic characteristic data of the road respectively; A multivariate Bayesian spatial model is constructed to assess the risk of road pedestrian collision based on the pedestrian data, road characteristic data, and traffic characteristic data.

2. The pedestrian collision risk micro-analysis method based on the multivariate Bayesian spatial model as claimed in claim 1, characterized in that: Extracting accident feature data includes: Obtain road pedestrian collision accident data, filter the accident data, exclude non-motor vehicle-pedestrian collision accidents, and only analyze motor vehicle-pedestrian collision accidents; Based on the road pedestrian collision accident data, information associated with the traffic accident is extracted as features, including the time of the accident, the location of the accident, age, gender, and weather conditions.

3. The pedestrian collision risk micro-analysis method based on the multivariate Bayesian spatial model as claimed in claim 1, characterized in that: Obtaining real-time pedestrian data on both directions of the road includes: Through real-time video stream, the YOLOv8 algorithm is used for target recognition to obtain pedestrian counting data in both directions of the road; Through time window division, the data is discretized into hourly data and daily data by time period, and then window aggregation is performed to count the number of pedestrians in each direction.

4. The pedestrian collision risk micro-analysis method based on a multivariate Bayesian spatial model as claimed in claim 3, characterized in that: The road feature data extracted from the road include: Obtain static road information through real-time video streams and extract road feature data of the road where the accident occurred; The road characteristic data includes road length, road width, road condition, road speed limit, bus stops and pedestrian crossings, wherein the road condition refers to road surface smoothness.

5. The pedestrian collision risk micro-analysis method based on the multivariate Bayesian spatial model as claimed in claim 4, characterized in that: The traffic characteristic data of the extracted roads include: The YOLOv8 vehicle detection model outputs bounding boxes for the video frames in the video data, assigns unique IDs in combination with the tracking algorithm, and counts the number of non-repeated IDs within the set time window to generate traffic flow data; The actual displacement is calculated based on the pixel displacement of the vehicle in consecutive video frames and the calibration scale, and the vehicle speed is obtained by dividing it by the frame interval time; According to the speed data of all vehicles, the maximum speed of the vehicle and the 85th percentile speed of the vehicle are obtained.

6. A pedestrian collision risk micro-analysis method based on a multivariate Bayesian spatial model as claimed in claim 1 or 5, characterized in that: Building a multivariate Bayesian spatial model involves: Poisson lognormal regression was constructed to establish the quantitative relationship between accident severity and road covariates. At the same time, the unobserved spatial heterogeneity and non-spatial heterogeneity were separated through spatial random effects and non-spatial random effects. The correlation between the counts of different injury severities was analyzed by multivariate model, and the random parameter was expressed as: ； in, represents a multivariate fixed parameter, represents a normally distributed random term.

7. The pedestrian collision risk micro-analysis method based on a multivariate Bayesian spatial model as claimed in claim 6, characterized in that: The Poisson lognormal regression is expressed as: ； in, It's on the road The observed incident severity levels were Number of pedestrian casualties, When represents minor injury, serious injury and death respectively, is the Poisson parameter, expressed as: ； ； in, represents the vector of covariates, is the intercept, is the parameter vector, represents an overdispersed unstructured random term, Structured random terms representing spatial dependencies, Indicates road segment i The severity of the accident k The total random effect term under .

8. A pedestrian collision risk micro-analysis system based on a multivariate Bayesian spatial model, using a pedestrian collision risk micro-analysis method based on a multivariate Bayesian spatial model as claimed in any one of claims 1 to 7, characterized in that: include, The accident data preprocessing module is used to collect road pedestrian collision accident data. Accidents are divided into three types: minor injuries, serious injuries and deaths. The accident data is screened and accident feature data is extracted. The pedestrian data module is used to detect the number of pedestrians in real time through monitoring data and obtain real-time pedestrian data in both directions of the road; A road traffic feature acquisition module is used to acquire road static information and road dynamic information, and to extract road feature data and traffic feature data of the road respectively; The pedestrian collision risk analysis module is used to construct a multivariate Bayesian spatial model to evaluate the road pedestrian collision risk based on the pedestrian data, road characteristic data and traffic characteristic data.

9. A computer device, characterized in that: include: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of a pedestrian collision risk micro-analysis method based on a multivariate Bayesian spatial model as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: It stores computer executable instructions, which, when executed by a processor, implement the steps of a pedestrian collision risk micro-analysis method based on a multivariate Bayesian spatial model as described in any one of claims 1 to 7.

Images