Road traffic potential safety hazard checking method based on railway crossing scene

Abstract

The invention discloses a road traffic potential safety hazard troubleshooting method based on a railway crossing scene. The method comprises the following steps: S1, constructing a railway crossing road traffic potential safety hazard troubleshooting knowledge system; s2, establishing a standardized data acquisition process; s3, the cleaning process is completed based on the checked road sections; s4, completing a stake number calibration process and completing a troubleshooting road section calibration process; s5, establishing an intelligent railway crossing image recognition method; s6, establishing a railway crossing road traffic potential safety hazard standardized troubleshooting method and summarization; by constructing the intelligent railway crossing hidden danger checking system, multi-angle data acquisition and processing are realized. The system integrates equipment such as a vehicle-mounted high-definition camera, a GPS (Global Positioning System) and a gyroscope, ensures synchronization and accuracy of acquired data, and accurately positions the geographic position of each image in combination with a stake mark calibration technology. The hidden danger types of traffic signs, marking lines and facilities are systematically analyzed for railway crossings with different guard modes.

Description

Technical Field

[0001] The present invention relates to the field of road traffic safety technology, and in particular to a method for detecting road traffic safety hazards based on a railway crossing scenario. Background Art

[0002] In recent years, my country has vigorously carried out road and railway infrastructure construction. Due to the limitations of terrain, geology, and railway or tram line construction factors, it is difficult to avoid the intersection of railroad tracks and roads in actual road construction, which easily leads to the hidden dangers of lack of traffic infrastructure and insufficient lighting facilities at railway crossings. Railway crossings are accident-prone points. When motor vehicles and pedestrians pass through railway crossings, they lack traffic language prompts, fail to stop, or collide due to poor visibility at night, resulting in a high accident rate and mortality rate at railway crossings.

[0003] The existing technology has the following deficiencies: Forming a standardized process for checking road traffic safety hazards at railway crossings. The existing railway crossing road traffic safety hazard inspection has not yet formed a corresponding standard process from data collection, inspection methods, inspection reports, result evaluation to inspection and management, and the railway crossing road traffic safety hazard inspection process lacks a theoretical basis.

[0004] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0005] The purpose of the present invention is to provide a method for detecting road traffic safety hazards based on a railway crossing scenario to solve the problems in the above-mentioned background technology.

[0006] In order to achieve the above object, the present invention provides the following technical solution: comprising the following steps: S1. Construct a knowledge system for checking road traffic safety hazards at railway crossings; S2. Establish a standardized data collection process; S3, completing the cleaning process based on the checked road sections; S4, completing the process of calibrating the pile number and completing the process of calibrating the road section; S5. Establish an intelligent railway crossing image recognition method; S6. Establish and summarize standardized methods for checking road traffic safety hazards at railway crossings.

[0007] Preferably, in S1, the specific steps are as follows: 1.1. Conduct in-depth analysis of the specific characteristics of railway crossings; the knowledge system constructed includes basic information about railway crossings, types of hidden dangers in traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities under different guarding methods, and covers hidden danger identification standards, location information of hidden danger inspection points, description of hidden danger content, and treatment suggestions, to ensure the comprehensiveness and accuracy of the inspection of road traffic safety hidden dangers at railway crossings; 1.2. Railway crossings are divided into two categories according to the guarding method: manned and unmanned. For manned railway crossings, the traffic signs to be checked include manned railway crossing signs, fork signs, speed limit signs and speed limit release signs; for unmanned railway crossings, the focus of the inspection is on stop and yield signs, speed limit signs, speed limit release signs and fork signs; for manned railway crossings, crossing barriers are set up, and the railway crossing as a whole should be equipped with lighting facilities, signal lights and video surveillance facilities.

[0008] Preferably, in S2, the specific steps are as follows: 2.1. Build a standardized data collection system based on intelligent equipment; the collection equipment includes vehicle-mounted high-definition cameras, GPS measuring instruments, gyroscopes, vehicle speed sensors and altitude measuring instruments; several high-definition cameras are arranged on the modified vehicles to collect multi-view and high-frequency road image data from the left front, front and right front; The vehicle's location information is collected through a GPS measuring instrument, the vehicle's roll angle, pitch angle and yaw angle are recorded using a gyroscope, the vehicle speed sensor measures the vehicle speed, and the altitude measuring instrument records the altitude during driving; 2.2. All data are stored in real time in the storage center of the device. The collected data are matched through time tags to ensure the synchronization of all data.

[0009] Preferably, in S3, the specific steps are as follows: According to the GPS information of the image, the collected road image data is mapped to the map; the original data is manually screened to retain the road section image data required for special inspections and determine the start and end positions of the inspection sections; during the cleaning process, the image data is mapped through GPS information and screened for duplicate removal; its core algorithm can be expressed as the following steps: Assume that the GPS information of the image data is expressed as longitude and latitude , which can be mapped to a specific location on the map according to the following formula: ,in, is the two-dimensional coordinate on the map, is the radius of the Earth, is the difference in longitude and latitude between two locations; Duplicate data cleaning algorithm: Since there may be duplicate images during the collection process, a hash value-based deduplication algorithm is required, which can be achieved through the following steps: Generate a unique hash value for each image ,in Represent images; compare hash values ​​of adjacent images: If they are the same, the image is marked as a duplicate and removed from the data set; data screening, during the data cleaning process, the screening conditions can be set manually to further filter the data; the screening conditions can be set by the following formula: ,in, is the image quality score, is the weight of each condition, are various evaluation functions, such as image clarity and timestamp.

[0010] Preferably, in S4, the specific steps are as follows: 4.1 During the inspection of the railway crossing, based on the information of the kilometer pile and the hundred-meter pile of the road, the present invention calibrates the electronic pile number on the system map; specifically, using the image data collected by the main camera and the right camera, the image containing the kilometer pile is found, and the system calibrates the pile number information based on the image of the right camera; according to the rule of setting a kilometer pile per kilometer and a hundred-meter pile per 100 meters, the system automatically calculates the electronic pile number information of each image to ensure that each image has clear location information; in addition, the system can automatically determine the uplink and downlink directions by the pile number; the distance is calculated by the GPS coordinates of the two pile number points, and the following formula can be used: ,in, is the actual distance between the two pile points, is the radius of the Earth, and The longitude and latitude of the two stakes are respectively; the electronic stake of the image is automatically calibrated using the rules of kilometer stakes and hundred-meter stakes; assuming that the starting stake is known to be ,,but: ,in: is the stake number corresponding to the image, is the distance from the image to the starting stake number. The system automatically calculates the electronic stake number information of each image based on the above calculation; 4.2 Based on the task creation information and image data, the system calibrates the name of the road section, the region to which it belongs, the jurisdiction unit, the up and down directions, the starting and ending pile numbers, the inspection mileage, the data collection time and the speed limit information; this process ensures the integrity and consistency of the inspection of road traffic safety hazards at railway crossings; for the calculation of the road section distance, it is assumed that the GPS coordinates of the starting and ending points are and , the distance formula can be used to calculate the length of the inspection section: ,Data calibration algorithm, each image of the road section to be checked needs to be calibrated to specific road section attributes, such as name, area, uplink and downlink directions, etc. The calibration can be achieved through the following algorithm: Get the GPS coordinates of each image; Match the image coordinates to the known road section information library: match After successful calibration, bind the image data to the road segment attributes: Road section name, up and down direction, stake number.

[0011] Preferably, in S5, the specific steps are as follows: when conducting the inspection of road traffic safety hazards at railway crossings, the layout of railway crossing traffic infrastructure is related to the location and type of the railway crossing, so it is necessary to realize railway crossing image recognition before the railway crossing inspection to clarify the type of railway crossing; railway crossings can be divided into manned railway crossings and unmanned railway crossings according to the guarding situation; Railway crossing recognition is mainly based on the currently very mature image recognition algorithm. The railway crossing image recognition algorithm is as follows: Step 1: Vehicle image input Input road collection of 4 images per second; Step 2: Get image information Obtain image information and perform grayscale processing on the image; Step 3: Image preprocessing Perform image field calculation, segmentation, equalization, smoothing, enhancement, binarization, and thinning techniques; Step 4: Image Recognition Extract rail features from pixel images obtained through image segmentation, and classify images based on whether they have rail features; Step 5: Image processing again Perform image preprocessing and image segmentation on images with railroad tracks; Step 6: Image Recognition Guardhouse features are extracted based on the pixel images obtained by image segmentation, and images are classified based on whether they have guardhouse features or not; Step7: Output the results.

[0012] Preferably, in S6, the specific steps are as follows: a standardized inspection method for road traffic safety hazards at railway crossings; the inspection method for road traffic safety hazards at railway crossings is based on the constructed knowledge system for inspection of safety hazards, and the inspection of road traffic safety hazards at railway crossings is carried out from different aspects such as traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities; The layout of railway crossing traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities is related to the road speed limit, railway crossing location, and type. After constructing the basic information of the railway crossing and clarifying the location and type of the railway crossing, the identification standards for the corresponding hidden dangers of the above-mentioned inspection contents can be clarified; the inspection method for road traffic safety hidden dangers at railway crossings relies on intelligent collection of pictures to conduct visual inspection of railway crossing traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities; the inspection steps are as follows: Step 1: According to the above scene recognition method, the inspection scene is determined to be a railway crossing. According to image recognition, the intersection of the outer edge of the outermost rail near the driving direction and the edge of the road is determined as the railway crossing as the reference point. ; Step 2: According to the standard provisions in the knowledge system of road traffic safety hazard inspection and the location information of safety hazards in railway crossings, determine the best points for inspection of traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities. ,in , The distance or range between the inspection points and the reference points of traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities specified in relevant standards; Step 3: Check from the driver's perspective and determine the best location The following picture is the best angle to check for road traffic safety hazards ; (1) Determine the best viewing angle for traffic signs: Check the cantilever traffic signs on the roadside. Due to the driver's field of vision, the best viewing angle Need to be at the best point A certain distance in front, according to ergonomics, the driver observes the cantilever column The upper angle of the traffic sign is 23°, and the distance between the driver and the ground is 1.2m; and Distance Spacing for: Check roadside column traffic signs ,Right now

[0013] Determine the best viewing angle for traffic markings: Check the road traffic markings. Due to the driver's field of vision, according to ergonomics, the lower limit angle for the driver to observe the road traffic markings is 8°, and the distance between the driver and the ground is 1.2m; and Distance Spacing 8.54m:

[0014] Determine the best viewing angle for technology facilities: Check the cantilever technology facilities on the roadside. Due to the driver's field of vision, the best viewing angle Need to be at the best point A certain distance in front, according to ergonomics, the driver observes the cantilever column The upper angle of the traffic sign is 23°, and the distance between the driver and the ground is 1.2m; and Distance Spacing for:

[0015] Indeed, the best view of other safety facilities: Crossing guardrails and lighting facilities ,Right now

[0016] Step 4: Confirm whether the road conditions of the railway crossing meet the regulatory requirements at the best viewing angle based on the image, and whether the traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities are missing at the best viewing angle: If there are no safety hazards at the best viewing angle, the hidden danger inspection for this point is completed; If the road conditions under the best viewing angle do not meet the regulatory requirements, and there are safety hazards in traffic signs, traffic markings, technological facilities, safety facilities, and lighting facilities, you need to go to Step 5 to confirm the details of traffic signs, traffic markings, technological facilities, safety facilities, and lighting facilities; Step 5: Manual adjustment The adjustment range is -1m to 1m, so as to confirm whether traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities are missing before and after the best inspection angle: If adjusted If no traffic signs, traffic markings, scientific and technological facilities, safety facilities, or lighting facilities are found after the inspection, they will be deemed to be missing; If adjusted If you find traffic signs, traffic markings, technological facilities, safety facilities, and lighting facilities after the value is reached, you need to go to Step 6 to determine the specific hidden dangers; Step 6: Determine whether traffic signs, traffic markings, technological facilities, safety facilities, and lighting facilities are complete: If they are complete, it is determined that the location of traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities do not meet the standards; If it is incomplete, combine it with Step 5 to comprehensively give the hidden danger identification results; Step 7: Output the hidden danger identification results in the form of picture annotations and text summaries to complete the intelligent investigation of safety hazards Step 8: Railway crossing road traffic safety hazard inspection process report, railway crossing safety hazard list, and overall report based on hazard inspection; the railway crossing road traffic safety hazard inspection process report can summarize the inspection results of railway crossing traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities.

[0017] In the above technical scheme, the technical effects and advantages provided by the present invention are as follows: by integrating on-board high-definition cameras, GPS measuring instruments, gyroscopes, speed sensors, altitude measuring instruments and other equipment, the system can collect high-frequency road images, location information and vehicle posture data from multiple angles, and ensure data synchronization through time tag matching technology. The system also removes redundant information through data cleaning, and accurately locates the geographical location of image data through pile number calibration to ensure the accuracy of the inspection information. In terms of hidden danger inspection, the present invention systematically analyzes the types of hidden dangers that may exist in traffic signs, markings, scientific and technological facilities, safety facilities, lighting facilities, etc. based on the guarding mode of railway crossings (manned and unmanned), and provides identification standards, location descriptions and rectification suggestions for hidden dangers. This not only improves the comprehensiveness and accuracy of hidden danger inspection, but also significantly improves the efficiency of safety management and reduces omissions in the manual inspection process.

[0018] The present invention realizes the standardization and intelligentization of hidden danger inspection at railway crossings, which can not only quickly identify and locate hidden dangers, but also effectively improve the efficiency of hidden danger inspection and treatment, reduce the incidence of traffic accidents, has important application value and economic benefits, and is suitable for the safety management and long-term maintenance of railway crossings. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0020] Figure 1 It is a schematic diagram of the technical route of the railway crossing traffic safety hazard investigation process of the present invention.

[0021] Figure 2 The figure is a flow chart of the railway crossing image recognition algorithm of the present invention.

[0022] Figure 3 It is a schematic diagram of the technical route of the method for troubleshooting road traffic safety hazards at railway crossings of the present invention. DETAILED DESCRIPTION

[0023] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.

[0024] The present invention provides Figures 1 to 3 A method for checking road traffic safety hazards based on a railway crossing scenario is shown, comprising the following steps: S1, constructing a knowledge system for checking road traffic safety hazards at railway crossings; S2, establishing a standardized data collection process; S3, completing a cleaning process based on the checked road sections; S4, completing the process of calibrating pile numbers and completing the process of calibrating the checked road sections; S5, establishing an intelligent railway crossing image recognition method; S6, establishing a method and summary for standardized checking of road traffic safety hazards at railway crossings.

[0025] In S1, the specific steps are as follows: 1.1, conduct an in-depth analysis of the specific characteristics of railway crossings; the knowledge system constructed includes the basic information of railway crossings, the types of hidden dangers existing in traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities under different guarding methods, and covers hidden danger identification standards, hidden danger inspection point location information, hidden danger content description and governance suggestions, to ensure the comprehensiveness and accuracy of the inspection of road traffic safety hidden dangers at railway crossings; railway crossings are special traffic scenes where roads and railways intersect, and the inspection of safety hidden dangers needs to be analyzed in depth in combination with the specific characteristics of railway crossings. In order to ensure the comprehensiveness and systematicness of the inspection, the present invention constructs a set of knowledge systems for the inspection of road traffic safety hidden dangers at railway crossings. The system is not only based on existing road traffic safety standards, but also covers unique hidden danger identification elements in railway crossing scenarios. The core part of the knowledge system includes comprehensive management of the basic information of railway crossings. The basic information covers key parameters such as road name, road speed limit, guarding method, number of motor vehicle lanes, and central isolation method. Through the standardized management of these basic information, the inspectors can quickly grasp the infrastructure of the railway crossings and provide strong support for subsequent hidden danger inspections. In addition, the system also lists in detail the types of potential hazards in traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities based on the different guarding methods of railway crossings. By combining hazard identification with the types and distribution of related facilities, inspectors can make accurate judgments based on the location information of hazard points.

[0026] In order to ensure the standardization and accuracy of the determination of hidden dangers, the knowledge system further covers the hidden danger identification standards. The hidden danger identification standards include the requirements for the normal working state of the facilities, the classification of hidden danger manifestations and the evaluation of the severity. The system will automatically compare the hidden danger points with the deviations of the facility functions to obtain the inspection results. In addition, the system also includes a detailed description of each hidden danger and treatment suggestions to help the inspection personnel quickly deal with and rectify the problems found, and ensure the continued safe operation of the railway crossing.

[0027] Through this knowledge system, the inspection of hidden dangers at railway crossings has been expanded from the identification of a single facility to a comprehensive analysis of the entire scenario, ensuring the integrity, standardization and accuracy of the inspection of road traffic safety hazards at railway crossings.

[0028] 1.2. Railway crossings are divided into two categories according to the guarding method: manned and unmanned. For manned railway crossings, the traffic signs to be checked include manned railway crossing signs, cross-shaped signs, speed limit signs, and speed limit release signs. For unmanned railway crossings, the focus is on checking stop and yield signs, speed limit signs, speed limit release signs, and cross-shaped signs. Manned railway crossings are equipped with crossing barriers, and the railway crossings as a whole should be equipped with lighting facilities, signal lights, and video surveillance facilities. Railway crossings can be divided into two categories according to the guarding method: manned and unmanned, and there are significant differences between the two in terms of traffic management methods and safety hazards. In order to ensure the effectiveness of the inspection work, the present invention conducts a detailed hidden danger inspection and analysis of the two types of railway crossings.

[0029] For manned railway crossings, traffic management relies on manual intervention and there are many types of facilities. The focus of hidden danger inspection is whether the facilities are fully equipped and whether their functions are functioning normally. The specific traffic signs to be inspected include the following categories: Manned Railroad Crossing Sign: This is the most basic sign used to alert road users to the existence of a railroad crossing and how it is managed. Check to see if the sign is legible, damaged, or missing.

[0030] Cross symbol (multiple railway tracks intersecting the road): used to alert drivers to the intersection of multiple railway tracks and the road. When checking, it is important to check the visibility of the symbol and whether it is equipped with auxiliary lighting in low light conditions.

[0031] Speed ​​limit signs and speed limit relief signs: used to set the speed of vehicles entering the railway crossing. When checking, you need to pay attention to whether the location of the sign is reasonable, whether the sign is blocked or damaged, and whether it is clearly visible under different weather conditions.

[0032] Crossing barriers: used to physically separate railways from roads, and are important facilities for guarding railway crossings. During inspection, it is necessary to check whether the opening and closing of the barriers are flexible, whether the control system operates normally, and whether it has protective functions.

[0033] For unmanned railway crossings, due to the lack of manual supervision, the safety hazards mainly rely on the monitoring and prompting of automated facilities. The traffic signs and equipment checked include: Stop and yield signs: used to remind drivers to stop and yield at railway crossings. During inspection, ensure that the signs are clearly visible, installed in the specified location, and not obscured by external influences (such as wind and rain).

[0034] Speed ​​limit signs and speed limit release signs: Similar to manned crossings, the inspection needs to check the integrity and correctness of the signs, especially in the special environment of unmanned railway crossings, the visibility and accuracy of the signs are more important.

[0035] Cross symbol: alerts drivers to the intersection of multiple railways. Check to ensure the symbol is conspicuous and visible at night or in low visibility conditions.

[0036] Whether it is a manned or unmanned railway crossing, lighting facilities, signal lights and video surveillance equipment should be installed to ensure that vehicle drivers can clearly identify the specific situation of the crossing at night or in bad weather conditions. During the inspection, it is necessary to confirm whether these devices are operating normally, whether the lighting is sufficient, whether the monitoring video images are clear and have no blind spots, and whether the color and brightness of the signal lights meet the specifications. Especially for unmanned railway crossings, the role of video surveillance is crucial, which can help remote managers to grasp the real-time situation of the railway crossing in a timely manner and respond quickly to potential safety hazards.

[0037] Through the detailed classification and description of the hidden danger inspection elements for manned and unmanned railway crossings, the present invention not only comprehensively covers the key safety hazards of different types of crossings, but also provides a clear inspection guide for inspectors to ensure the accuracy and operability of hidden danger inspection.

[0038] In S2, the specific steps are as follows: 2.1. Build a standardized data collection system based on intelligent equipment; the collection equipment includes vehicle-mounted high-definition cameras, GPS measuring instruments, gyroscopes, vehicle speed sensors and altitude measuring instruments; several high-definition cameras are arranged on the modified vehicles to collect multi-view and high-frequency road image data from the left front, front and right front; The GPS measuring instrument is used to collect the vehicle's driving position information, the gyroscope is used to record the vehicle's roll angle, pitch angle and yaw angle, the vehicle speed sensor is used to measure the vehicle speed, and the altitude measuring instrument is used to record the altitude during driving; the core of the system is to obtain real-time data of roads and railway crossings through a variety of high-precision sensing equipment. Specifically, the collection equipment includes the following key components: On-board high-definition camera: Multiple high-definition cameras are installed on the left front, front and right front of the modified vehicle to collect multi-angle and high-frequency road image data. These cameras can not only capture the panoramic view of the road, but also obtain detailed information such as the signs, markings, facilities, etc. of the railway crossing. Cameras with different perspectives work together to ensure that no potential safety hazards are missed during the inspection process. GPS measuring instrument: Real-time recording of the vehicle's driving position information, through the high-precision GPS system, can accurately determine the vehicle's position coordinates at every moment on the road. GPS data provides the location information basis for image data, so that each image has clear geographic coordinates. Gyroscope: used to record the changes in the vehicle's posture during driving, including the vehicle's roll angle (tilt around the front and rear axes), pitch angle (up and down tilt around the left and right axes) and yaw angle (rotation angle around the vertical axis). This information can reflect the vehicle's motion state under complex terrain or different road conditions, ensuring that the collected data is still accurate under different road conditions. Vehicle speed sensor: real-time monitoring of the vehicle's driving speed. Speed ​​information is crucial to the frequency and accuracy of image acquisition, ensuring the continuity and clarity of images during high-speed driving. Altitude meter: records the altitude of the vehicle during driving, especially at railway crossings and nearby mountains or undulating sections. Altitude data can provide additional terrain references to help analyze the slope of the road and the safety of the railway crossing. Through the coordinated work of multiple devices, the system can comprehensively monitor the conditions of the railway crossing and its surrounding roads, and ensure that the collected data is accurate and reliable, which can provide basic data for subsequent hidden danger inspections.

[0039] 2.2. All data are stored in the storage center of the device in real time. The collected data are matched through time tags to ensure the synchronization of various data. During the data collection process, all information collected by sensors and cameras will be transmitted and stored in the storage center of the device in real time to ensure the immediacy and continuity of the data. The storage center will process each type of data separately, including image data, GPS data, speed information, vehicle posture data and altitude information. In order to ensure the synchronization and consistency between different types of data, the system introduces a time tag matching mechanism. When each data is collected, a timestamp accurate to milliseconds will be attached. The time tag can uniformly mark data points from different sources to ensure the consistency of the collected data in the time dimension. This mechanism enables image data to correspond one by one with the specific location information, driving speed, posture change and altitude of the vehicle. For example, when the camera captures a road image at a certain moment, the system will automatically associate this image with the GPS location information, vehicle speed and altitude data at the corresponding moment according to the time tag. The advantage of this is that when inspectors view each image, they can not only understand the conditions of the roads and railway crossings shown in the image, but also simultaneously obtain the posture, speed and altitude information of the vehicle at that location, forming a multi-dimensional understanding of the road and railway crossing environment. In addition, the data storage center also has a data redundancy detection function, which can automatically identify and clean up redundant data collected repeatedly, ensuring that the data in the system is efficient and streamlined, and reducing the complexity of subsequent data processing. After each data collection is completed, the system will automatically organize and analyze the data, providing structured and visual data support for the subsequent railway crossing hidden danger inspection.

[0040] Through this standardized data collection and processing process, the system can achieve all-round, multi-angle monitoring of railway crossings and their surrounding road environments, providing stable technical support for hidden danger detection.

[0041] In S3, the specific steps are as follows: According to the GPS information of the image, the collected road image data is mapped to the map; the original data is manually screened to retain the road section image data required for special inspections and determine the start and end positions of the inspection sections; during the cleaning process, the image data is mapped through GPS information and screened for duplicate removal; its core algorithm can be expressed as the following steps: Assume that the GPS information of the image data is expressed as longitude and latitude , which can be mapped to a specific location on the map according to the following formula: ,in, is the two-dimensional coordinate on the map, is the radius of the Earth, is the difference in longitude and latitude between two locations; Duplicate data cleaning algorithm: Since there may be duplicate images during the collection process, a hash value-based deduplication algorithm is required, which can be achieved through the following steps: Generate a unique hash value for each image ,in Represent images; compare hash values ​​of adjacent images: If they are the same, the image is marked as a duplicate and removed from the data set; data screening, during the data cleaning process, the screening conditions can be set manually to further filter the data; the screening conditions can be set by the following formula: ,in, is the image quality score, is the weight of each condition, is a function of various evaluations, such as image clarity and timestamp; in the railway crossing scenario, data cleaning is a key step to ensure that the collected road image data is efficient and unique. The process first maps the image data to the map through GPS information so that the geographical location of the image collection can be visually viewed. Since there may be duplication or redundancy in the process of road data collection, the system will automatically identify and remove duplicate image data by analyzing the characteristics of each image to ensure that the image data of each road is unique; the process also includes further screening of image data, based on evaluation criteria in different dimensions such as road conditions, ambient light, and camera angles, to exclude images with poor quality or insufficient information. The cleaned data will provide a reliable basis for hidden danger inspection and ensure the accuracy and comprehensiveness of the inspection results. In addition, manual participation is introduced in the data cleaning process. For some complex sections or specific inspection needs, manual judgment is used to further improve the accuracy of image screening and optimize the intelligent decision-making of the system.

[0042] In S4, the specific steps are as follows: 4.1 During the inspection of the railway crossing, based on the information of the kilometer pile and the hundred-meter pile of the road, the present invention calibrates the electronic pile number on the system map; specifically, using the image data collected by the main camera and the right camera, the image containing the kilometer pile is found, and the system calibrates the pile number information based on the image of the right camera; according to the rule of setting a kilometer pile per kilometer and a hundred-meter pile per 100 meters, the system automatically calculates the electronic pile number information of each image to ensure that each image has clear location information; in addition, the system can automatically determine the uplink and downlink directions through the pile number; the distance is calculated by the GPS coordinates of the two pile number points, and the following formula can be used: ,in, is the actual distance between the two pile points, is the radius of the Earth, and are the longitude and latitude of the two stakes respectively; assuming that the starting stake is known to be ,but: ,in: is the stake number corresponding to the image, is the distance from the image to the starting stake number. The system automatically calculates the electronic stake number information of each image based on the above calculation; 4.2 Based on the task creation information and image data, the system calibrates the name of the road section, the region to which it belongs, the jurisdiction unit, the up and down directions, the starting and ending pile numbers, the inspection mileage, the data collection time and the speed limit information; this process ensures the integrity and consistency of the inspection of road traffic safety hazards at railway crossings; for the calculation of the road section distance, it is assumed that the GPS coordinates of the starting and ending points are and , the distance formula can be used to calculate the length of the inspection section: ,Data calibration algorithm, each image of the road section to be checked needs to be calibrated to specific road section attributes, such as name, area, uplink and downlink directions, etc. The calibration can be achieved through the following algorithm: Get the GPS coordinates of each image; Match the image coordinates to the known road section information library: match After successful calibration, bind the image data to the road segment attributes: Road section name, up and down direction, stake number; stake number calibration is an important step to ensure the position accuracy of image data in the inspection of railway crossing safety hazards. By calibrating the kilometer stakes and hundred-meter stakes of each section of road, the system can give accurate location information to each collected image. In actual operation, multiple cameras of the front-end acquisition equipment will capture images of kilometer stakes and hundred-meter stakes on the right side of the road. The system automatically identifies these stake numbers and associates them with the corresponding image data.

[0043] When a kilometer pile or a hundred-meter pile is identified, the system will automatically calculate the pile number information of all images based on the distribution rules of the pile numbers on the road. Each pile number information not only represents the specific location of the image, but also includes the determination of the uplink and downlink directions. Through this precise calibration method, the system can achieve comprehensive monitoring of each section of the road, ensuring that each image has clear geographic coordinates and driving direction, which facilitates the subsequent accurate analysis and processing of potential danger points.

[0044] In S5, the specific steps are as follows: when conducting the inspection of road traffic safety hazards at railway crossings, the layout of railway crossing traffic infrastructure is related to the location and type of railway crossings. Therefore, before the railway crossing inspection, it is necessary to realize railway crossing image recognition and clarify the type of railway crossing; railway crossings can be divided into manned railway crossings and unmanned railway crossings according to the guarding situation; Railway crossing recognition is mainly based on the currently very mature image recognition algorithm, among which the railway crossing image recognition algorithm is as follows: Step 1: Vehicle image input Input road collection of 4 images per second; Step 2: Get image information Obtain image information and perform grayscale processing on the image; Step 3: Image preprocessing Perform image field calculation, segmentation, equalization, smoothing, enhancement, binarization, and thinning techniques; Step 4: Image Recognition Extract rail features from pixel images obtained through image segmentation, and classify images based on whether they have rail features; Step 5: Image processing again Perform image preprocessing and image segmentation on images with railroad tracks; Step 6: Image Recognition Guardhouse features are extracted based on the pixel images obtained by image segmentation, and images are classified based on whether they have guardhouse features or not; Step 7: Output the results. Clarify the standardized inspection method for road traffic safety hazards at railway crossings. The inspection method for road traffic safety hazards at railway crossings is based on the constructed knowledge system for inspection of safety hazards, and conducts inspection of road traffic safety hazards at railway crossings from different aspects such as traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities.

[0045] In S6, the specific steps are as follows: Standardized inspection method for road traffic safety hazards at railway crossings; The inspection method for road traffic safety hazards at railway crossings is based on the constructed knowledge system for inspection of safety hazards, and the inspection of road traffic safety hazards at railway crossings is carried out from different aspects such as traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities; The layout of railway crossing traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities is related to the road speed limit, railway crossing location, and type. After constructing the basic information of the railway crossing and clarifying the location and type of the railway crossing, the identification standards for the corresponding hidden dangers of the above-mentioned inspection contents are clear; the inspection method for road traffic safety hidden dangers at railway crossings relies on intelligent collection of pictures to conduct visual inspection of railway crossing traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities; the inspection steps are as follows: Step 1: According to the above scene recognition method, the inspection scene is determined to be a railway crossing. According to image recognition, the intersection point between the outer edge of the outermost rail near the driving direction and the edge of the road is determined as - the railway crossing is determined as the reference point ; Step 2: According to the standard provisions in the knowledge system of road traffic safety hazard inspection and the location information of safety hazards in railway crossings, determine the best points for inspection of traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities. ,in , The distance or range between the inspection points and the reference points of traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities specified in relevant standards; Step 3: Check from the driver's perspective and determine the best location The picture below is the best angle to check for road traffic safety hazards; (1) Determine the best viewing angle for traffic signs: Check the cantilever traffic signs on the roadside. Due to the driver's field of vision, the best viewing angle Need to be at the best point A certain distance in front, according to ergonomics, the driver observes the cantilever column The upper angle of the traffic sign is 23°, and the distance between the driver and the ground is 1.2m; and Distance Spacing for:

[0046] Check roadside column traffic signs ,Right now

[0047] Determine the best viewing angle for traffic markings: Check the road traffic markings. Due to the driver's field of vision, according to ergonomics, the lower limit angle for the driver to observe the road traffic markings is 8°, and the distance between the driver and the ground is 1.2m; and Distance Spacing 8.54m:

[0048] Determine the best viewing angle for technology facilities: Check the cantilever technology facilities on the roadside. Due to the driver's field of vision, the best viewing angle Need to be at the best point A certain distance in front, according to ergonomics, the driver observes the cantilever column The upper angle of the traffic sign is 23°, and the distance between the driver and the ground is 1.2m; and Distance Spacing for:

[0049] Indeed, the best view of other safety facilities: Crossing guardrails and lighting facilities ,Right now

[0050] Step 4: Confirm whether the road conditions of the railway crossing meet the regulatory requirements at the best viewing angle based on the image, and whether the traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities are missing at the best viewing angle: If there are no safety hazards at the best viewing angle, the hidden danger inspection for this point is completed; If the road conditions under the best viewing angle do not meet the regulatory requirements, and there are safety hazards in traffic signs, traffic markings, technological facilities, safety facilities, and lighting facilities, you need to go to Step 5 to confirm the details of traffic signs, traffic markings, technological facilities, safety facilities, and lighting facilities; Step 5: Manual adjustment The adjustment range is -1m to 1m, so as to confirm whether traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities are missing before and after the best inspection angle: If adjusted If no traffic signs, traffic markings, scientific and technological facilities, safety facilities, or lighting facilities are found after the inspection, they will be deemed to be missing; If adjusted If you find traffic signs, traffic markings, technological facilities, safety facilities, and lighting facilities after the value is reached, you need to go to Step 6 to determine the specific hidden dangers; Step 6: Determine whether traffic signs, traffic markings, technological facilities, safety facilities, and lighting facilities are complete: If it is complete, it is determined that the location of traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities do not meet the standards. If it is incomplete, combine it with Step 5 to comprehensively give the hidden danger identification results; Step 7: Output the hidden danger identification results in the form of picture annotations and text summaries to complete the intelligent investigation of safety hazards Step 8: Railway crossing road traffic safety hazard inspection process report, railway crossing safety hazard list, and overall report based on hazard inspection; intelligent inspection of railway crossing road traffic safety hazards combines intelligent algorithms with manual assistance to determine the best viewing angle for railway crossing hazard inspection, selects image-related data of target points for hazard inspection, first identifies hazards using deep image recognition technology, and assists with manual judgment to ensure a comprehensive inspection of railway crossing road traffic safety hazards, and marks the hazards found in the inspection in the form of image tags, so that the hazard-marked images can be attached to the inspection report, and the inspection results can be displayed in a visual form. The railway crossing road traffic safety hazard inspection process report can summarize the inspection results of railway crossing traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities. In order to achieve hazard inspection conclusions for different users, a hierarchical inspection report is defined: including railway crossing road traffic safety hazard inspection process report, railway crossing safety hazard list, and overall report based on hazard inspection. The railway crossing road traffic safety hazard inspection process report can summarize the inspection results of railway crossing traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities. Its role is to ensure the integrity and correctness of the inspection process; the railway crossing safety hazard list and the overall report based on the hazard inspection analyze the existing hazard content from part to whole. Its role is to view the hazard inspection conclusions in different permissions and scopes according to different user business needs. The above-defined inspection report is based on the inspection report format requirements in the inspection work specifications, which is convenient for reporting the inspection conclusions to relevant departments during the road traffic safety hazard inspection work and for the subsequent safety hazard management.

[0051] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for checking potential road traffic safety hazards based on a railway crossing scenario, characterized by: The following steps are involved: S1. Construct a knowledge system for checking road traffic safety hazards at railway crossings; S2. Establish a standardized data collection process; S3, completing the cleaning process based on the checked road sections; S4, completing the process of calibrating the pile number and completing the process of calibrating the road section; S5. Establish an intelligent railway crossing image recognition method; S6. Establish and summarize standardized methods for checking road traffic safety hazards at railway crossings.

2. According to claim 1, a method for troubleshooting road traffic safety hazards based on a railway crossing scenario is characterized by: In S1, the specific steps are as follows: 2.

1. Conduct in-depth analysis of the specific characteristics of railway crossings; the knowledge system constructed includes basic information about railway crossings, types of hidden dangers in traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities under different guarding methods, and covers hidden danger identification standards, location information of hidden danger inspection points, description of hidden danger content, and treatment suggestions, to ensure the comprehensiveness and accuracy of the inspection of road traffic safety hidden dangers at railway crossings; 2.

2. Railway crossings are divided into two categories based on the guarding method: manned and unmanned. For manned railway crossings, the traffic signs to be checked include manned railway crossing signs, fork signs, speed limit signs and speed limit release signs. For unmanned railway crossings, the focus of the inspection is on stop and yield signs, speed limit signs, speed limit release signs and fork signs. For manned railway crossings, crossing guardrails are installed, and the railway crossing as a whole should be equipped with lighting facilities, signal lights and video surveillance facilities.

3. The method for troubleshooting road traffic safety hazards based on a railway crossing scenario according to claim 1 is characterized by: In S2, the specific steps are as follows: 3.

1. Build a standardized data collection system based on intelligent equipment; the collection equipment includes vehicle-mounted high-definition cameras, GPS measuring instruments, gyroscopes, vehicle speed sensors and altitude measuring instruments; several high-definition cameras are arranged on the modified vehicles to collect multi-view and high-frequency road image data from the left front, front and right front; The vehicle's location information is collected through a GPS measuring instrument, the vehicle's roll angle, pitch angle and yaw angle are recorded using a gyroscope, the vehicle speed sensor measures the vehicle speed, and the altitude measuring instrument records the altitude during driving; 3.

2. All data are stored in real time in the storage center of the device. The collected data are matched through time tags to ensure the synchronization of all data.

4. The method for troubleshooting road traffic safety hazards based on a railway crossing scenario according to claim 1 is characterized by: In S3, the specific steps are as follows: According to the GPS information of the image, the collected road image data is mapped onto the map; the original data is manually screened to retain the road section image data required for the special inspection and determine the start and end positions of the inspection section; during the cleaning process, the image data is mapped through the GPS information and screened to remove duplicates; Its core algorithm can be expressed as the following steps: Assume that the GPS information of the image data is expressed as longitude and latitude , which can be mapped to a specific location on the map according to the following formula: ,in, is the two-dimensional coordinate on the map, is the radius of the Earth, is the difference in longitude and latitude between two locations; Duplicate data cleaning algorithm: Since there may be duplicate images during the collection process, a hash value-based deduplication algorithm is required, which can be achieved through the following steps: Generate a unique hash value for each image ,in Represent images; compare hash values ​​of adjacent images: If they are the same, the image is marked as a duplicate and removed from the data set; data screening, during the data cleaning process, the screening conditions can be set manually to further filter the data; the screening conditions can be set by the following formula: ,in, is the image quality score, is the weight of each condition, are various evaluation functions, such as image clarity and timestamp.

5. The method for troubleshooting road traffic safety hazards based on a railway crossing scenario according to claim 1 is characterized by: In S4, the specific steps are as follows: 5.1 During the inspection of railway crossings, the present invention calibrates the electronic stake number on the system map based on the information of the kilometer stake and the hundred-meter stake of the road; specifically, the image data collected by the main camera and the right camera is used to find the image containing the kilometer stake, and the system calibrates the stake number information based on the image of the right camera; according to the rule of setting a kilometer stake for every kilometer and a hundred-meter stake for every 100 meters, the system automatically calculates the electronic stake number information of each image to ensure that each image has clear location information; in addition, the system can automatically determine the uplink and downlink directions through the stake number; the distance can be calculated through the GPS coordinates of two stake number points, and the following formula can be used: ,in, is the actual distance between the two pile points, is the radius of the Earth, and are the latitude and longitude of the two stakes respectively; assuming that the starting stake is known to be ,but: ,in: is the stake number corresponding to the image, is the distance from the image to the starting stake number. The system automatically calculates the electronic stake number information of each image based on the above calculation; 5.2 Based on the task creation information and image data, the system calibrates the name of the road section, the region to which it belongs, the jurisdiction unit, the up and down directions, the starting and ending pile numbers, the inspection mileage, the data collection time and the speed limit information; this process ensures the integrity and consistency of the inspection of road traffic safety hazards at railway crossings; for the calculation of the road section distance, it is assumed that the GPS coordinates of the starting and ending points are and , the distance formula can be used to calculate the length of the inspection section: ,Data calibration algorithm, each image of the road section to be checked needs to be calibrated to a specific road section attribute, which can be achieved through the following calibration algorithm: ,Get the GPS coordinates of each image; Match the image coordinates to the known road section information database: match After successful calibration, bind the image data to the road segment attributes: It is the road section name, up and down direction, and pile number.

6. The method for troubleshooting road traffic safety hazards based on a railway crossing scenario according to claim 1 is characterized by: In S5, the specific steps are as follows: when conducting the inspection of road traffic safety hazards at railway crossings, the layout of railway crossing traffic infrastructure is related to the location and type of railway crossings. Therefore, before the railway crossing inspection, it is necessary to realize railway crossing image recognition and clarify the type of railway crossing; railway crossings can be divided into manned railway crossings and unmanned railway crossings according to the guarding situation; Railway crossing recognition is mainly based on the currently very mature image recognition algorithm. The railway crossing image recognition algorithm is as follows: Step 1: Vehicle image input Input road collection of 4 images per second; Step 2: Get image information Obtain image information and perform grayscale processing on the image; Step 3: Image preprocessing Perform image field calculation, segmentation, equalization, smoothing, enhancement, binarization, and thinning techniques; Step 4: Image Recognition Extract rail features from pixel images obtained through image segmentation, and classify images based on whether they have rail features; Step 5: Image processing again Perform image preprocessing and image segmentation on images with railroad tracks; Step 6: Image Recognition Guardhouse features are extracted based on the pixel images obtained by image segmentation, and images are classified based on whether they have guardhouse features or not; Step7: Output the results.

7. The method for troubleshooting road traffic safety hazards based on a railway crossing scenario according to claim 1 is characterized by: In S6, a standardized inspection method for road traffic safety hazards at railway crossings is developed; the inspection method for road traffic safety hazards at railway crossings is based on the established knowledge system for inspection of safety hazards, and the inspection of road traffic safety hazards at railway crossings is carried out from different aspects such as traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities; The layout of railway crossing traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities is related to the road speed limit, railway crossing location, and type. After constructing the basic information of the railway crossing and clarifying the location and type of the railway crossing, the identification standards for the corresponding hidden dangers of the above-mentioned inspection contents can be clarified; the inspection method for road traffic safety hidden dangers at railway crossings relies on intelligent collection of pictures to conduct visual inspection of railway crossing traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities; the inspection steps are as follows: Step 1: According to the above scene recognition method, the inspection scene is determined to be a railway crossing. According to image recognition, the intersection point between the outer edge of the outermost rail near the driving direction and the edge of the road is determined as the railway crossing as the reference point. ; Step 2: According to the standard provisions in the knowledge system of road traffic safety hazard inspection and the location information of safety hazards in railway crossings, determine the best points for inspection of traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities. ,in, , The distance and range between the inspection points and the reference points of traffic signs, traffic markings, scientific and technological facilities, safety facilities and lighting facilities specified in relevant standards; Step 3: Check from the driver's perspective and determine the best location The following picture is the best angle to check for road traffic safety hazards ; Determine the best viewing angle for traffic signs: Check the cantilever traffic signs on the roadside. Due to the driver's field of vision, the best viewing angle Need to be at the best point A certain distance in front, according to ergonomics, the driver observes the cantilever column The upper angle of the traffic sign is 23°, and the distance between the driver and the ground is 1.2m; and Distance Spacing for: ； Check roadside column traffic signs ,Right now ; Determine the best viewing angle for traffic markings: Check the road traffic markings. Due to the driver's field of vision, according to ergonomics, the lower limit angle for the driver to observe the road traffic markings is 8°, and the distance between the driver and the ground is 1.2m; and Distance Spacing 8.54m: ； Determine the best viewing angle for technology facilities: Check the cantilever technology facilities on the roadside. Due to the driver's field of vision, the best viewing angle Need to be at the best point A certain distance in front, according to ergonomics, the driver observes the cantilever column The upper angle of the traffic sign is 23°, and the distance between the driver and the ground is 1.2m; and Distance Spacing for: ； Indeed, the best view of other safety facilities: Crossing guardrails and lighting facilities ,Right now ; Step 4: Confirm whether the road conditions of the railway crossing meet the regulatory requirements at the best viewing angle based on the image, and whether the traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities are missing at the best viewing angle: If there are no safety hazards at the best viewing angle, the hidden danger inspection for this point is completed; If the road conditions under the best viewing angle do not meet the regulatory requirements, and there are safety hazards in traffic signs, traffic markings, technological facilities, safety facilities, and lighting facilities, you need to go to Step 5 to confirm the details of traffic signs, traffic markings, technological facilities, safety facilities, and lighting facilities; Step 5: Manual adjustment The adjustment range is -1m to 1m, so as to confirm whether traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities are missing before and after the best inspection angle: If adjusted If no traffic signs, traffic markings, scientific and technological facilities, safety facilities, or lighting facilities are found after the inspection, they will be deemed to be missing; If adjusted If you find traffic signs, traffic markings, technological facilities, safety facilities, and lighting facilities after the value is reached, you need to go to Step 6 to determine the specific hidden dangers; Step 6: Determine whether traffic signs, traffic markings, technological facilities, safety facilities, and lighting facilities are complete: If it is complete, it is determined that the location of traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities do not meet the standards. If it is incomplete, combine it with Step 5 to comprehensively give the hidden danger identification results; Step 7: Output the hidden danger identification results in the form of picture annotations and text summaries to complete the intelligent investigation of safety hazards Step 8: Railway crossing road traffic safety hazard inspection process report, railway crossing safety hazard list, and overall report based on hazard inspection; the railway crossing road traffic safety hazard inspection process report can summarize the inspection results of railway crossing traffic signs, traffic markings, scientific and technological facilities, safety facilities, and lighting facilities.

Images