Traffic accident reconstruction method and device based on automobile event data recording system

Abstract

The invention discloses a traffic accident reconstruction method and device based on an automobile event data recording system. The method comprises the following steps: acquiring field investigation data information, monitoring video image information and EDR data of an accident vehicle after a traffic accident field; performing data fusion complementation on the scene investigation data information, the monitoring video image information and the verified EDR data of the accident vehicle to obtain a data fusion result; and performing automatic modeling based on the data fusion result to construct a standardized accident reconstruction process to obtain an accident reconstruction result, and verifying the accident reconstruction result. According to the method, the working condition information of the vehicle in the accident process is efficiently and accurately obtained by adopting a more scientific method, and the efficiency and accuracy of accident reproduction can be remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of accident reconstruction, and in particular to a traffic accident reconstruction method and device based on an automobile event data recording system. Background Art

[0002] With the rapid increase in the number of motor vehicles on the road, the incidence of traffic accidents has also risen, posing a huge threat to the safety of public life and property. In order to effectively prevent and reduce such accidents, it is particularly important to deeply understand and analyze the causes of accidents. Traditional accident reconstruction mainly relies on manual on-site investigations and descriptions of the parties involved in the accident. This method has limitations in efficiency and accuracy and is easily affected by personal subjective factors. However, with the widespread use of automobile event data recording systems (EDR), we have obtained new technical means to reconstruct traffic accidents.

[0003] The existing accident reconstruction methods basically include two technical bases: one is the reconstruction work based on the data and scenes obtained from the traditional on-site investigation; the other is the traffic accident reconstruction method based on surveillance video images. (1) The data obtained by traditional on-site investigation methods are usually limited to the damage to the vehicle at the scene after the accident, scattered objects on the road, road traces, damage to road facilities and buildings, etc. These data cannot directly reveal the situation before the accident, such as the vehicle's driving trajectory, speed, acceleration and deceleration, and braking. Technicians need to analyze and calculate based on the investigation data to infer the state before the accident. This process not only tests the technical ability of the technicians themselves, but also depends greatly on the accuracy of the investigation data. Therefore, this practice makes it difficult to standardize the accident reconstruction work and has a weak scientific continuity. (2) Although surveillance videos can reflect the accident process, they usually have limitations. The videos taken by fixed road probes may not be able to accurately and clearly reproduce the accident process due to factors such as shooting angle, distance, and pixels, and are therefore difficult to use for accurate reconstruction. On-board surveillance videos usually only record the image information in front of the vehicle and often cannot fully display collision accidents occurring on the side or rear of the vehicle. Therefore, image-based reconstruction methods cannot accurately reflect the operation of the occupants, especially for vehicles with technical problems such as braking and steering. (3) Existing on-site reconstruction methods have not fully covered the working conditions of the passive safety systems in the car, such as airbags, seat belts, etc., as well as some active safety systems, such as automatic emergency braking systems, anti-lock braking systems, and body stability control systems during the accident. This limitation may make it impossible to accurately restore the accident process, especially the injury process of the occupants. Summary of the invention

[0004] The present invention aims to solve one of the technical problems in the related art at least to a certain extent.

[0005] To this end, the present invention proposes a traffic accident reconstruction method based on the automobile event data recording system and an EDR-based traffic accident reconstruction platform, which, on the basis of traditional survey data, takes EDR data as the core to achieve accurate, efficient, scientific and objective reconstruction of the accident process.

[0006] Another object of the present invention is to provide a traffic accident reconstruction device based on an automobile event data recording system.

[0007] To achieve the above-mentioned object, the present invention provides a traffic accident reconstruction method based on a vehicle event data recording system, comprising:

[0008] Obtain on-site investigation data information, surveillance video image information and EDR data of the accident vehicle after the traffic accident scene;

[0009] Performing data fusion and completion on the on-site investigation data information, the monitoring video image information, and the verified EDR data of the accident vehicle to obtain a data fusion result;

[0010] Based on the data fusion results, automated modeling is performed to construct a standardized accident reconstruction process to obtain accident reconstruction results, and the accident reconstruction results are verified.

[0011] The traffic accident reconstruction method based on the automobile event data recording system of the embodiment of the present invention may also have the following additional technical features:

[0012] In one embodiment of the present invention, the on-site investigation data information includes: static space information and accident-related information; the monitoring video image information includes basic image information and information reflecting the accident; and the EDR data includes basic information and working condition information.

[0013] In one embodiment of the present invention, the static spatial information includes road surface spatial dimensions, road signs and markings, road construction information, and road surface adhesion coefficient;

[0014] The accident-related information includes driver information, driving and passenger relationships in the vehicle, signs of vehicle damage, and marks left on the road surface;

[0015] The basic information of the image, including name and size, format and frame rate, integrity check value, and whether it is recorded clearly;

[0016] The accident information includes the vehicle's driving trajectory, the vehicle's collision process, braking before the collision, and steering before the collision;

[0017] The basic information includes basic vehicle information, system status during reading, software and hardware used for reading, data record completeness, data source description and data limitations;

[0018] The operating condition information includes driving speed, longitudinal acceleration change, lateral acceleration change, collision type, vehicle braking condition, vehicle steering condition, vehicle acceleration condition, engine speed, seat belt and airbag working conditions, ABS and AEB working conditions, and ACC and LDW working conditions.

[0019] In one embodiment of the present invention, the EDR data of the accident vehicle is verified, including: relevance verification, integrity verification and validity verification; wherein,

[0020] The correlation verification is as follows: verification of the consistency between the vehicle identification code recorded in the EDR data and the accident vehicle identification code, so as to obtain the consistency result of the vehicle identification code; verification that the power-on cycle at the time of the accident recorded in the EDR data is similar to the power-on cycle when the EDR data is read, so as to obtain the consistency verification result of the power-on cycle; verification that the airbag deployment recorded in the EDR data is consistent with the airbag detonation period shown in the accident vehicle inspection photos and the on-board video, so as to obtain the consistency verification result of the airbag deployment; verification that the two collision forms and time intervals recorded in the EDR data are consistent with those shown in the accident vehicle inspection photos, on-board videos, and on-site photos, so as to obtain the collision form and time interval consistency verification result;

[0021] The integrity verification: verifying that the event data recorded in the EDR data should all be represented as successfully and completely recorded, so as to obtain the data record integrity verification result;

[0022] The validity verification: performs tire specification verification, collision analysis and time node analysis to obtain validity verification results.

[0023] In one embodiment of the present invention, automatic modeling is performed based on the data fusion result to construct a standardized accident reconstruction process to obtain an accident reconstruction result, including:

[0024] Use OCR technology and natural language processing technology to process on-site investigation data information to automatically identify and extract text information in a targeted manner, and perform structured processing; use computer vision technology to process surveillance video image information, and parse EDR data to convert it into a format recognizable by the PC-crash model to obtain pre-processed data; fuse the pre-processed data to obtain a complete accident data set;

[0025] Establish vehicle model library and scene model library, and dynamically update and manage them; automatically match the corresponding vehicle model and scene model according to the accident data set, and set parameters; run PC-crash simulation calculations to generate accident reconstruction animations;

[0026] Analyze the accident reconstruction animation, obtain relevant parameters of the accident process, and visualize the accident reconstruction results; compare the accident reconstruction results with investigation traces, surveillance videos and EDR data to verify the accuracy of the accident reconstruction results; automatically generate an accident reconstruction report.

[0027] To achieve the above-mentioned purpose, a second embodiment of the present invention provides a traffic accident reconstruction device based on a vehicle event data recording system, comprising:

[0028] The data acquisition module is used to obtain on-site investigation data information, monitoring video image information and EDR data of the accident vehicle after the traffic accident scene;

[0029] A data judgment module is used to judge whether the EDR data has relevance, integrity and validity according to the on-site investigation data information and the monitoring video image information to obtain a data judgment result;

[0030] A data fusion module is used to fuse the EDR data with the on-site investigation data information and the monitoring video image information based on the data judgment result to obtain a data fusion result;

[0031] The accident reconstruction module is used to perform accident reconstruction according to the data fusion result to obtain a reconstruction result.

[0032] The traffic accident reconstruction method and device based on the automobile event data recording system of the embodiment of the present invention can, firstly, automatically extract and analyze the data in the EDR, quickly reproduce the accident process, thereby improving the accuracy and efficiency of reconstruction, reducing the differences between manual intervention and subjective judgment, enhancing standardization, and reducing the impact caused by differences in technical personnel's experience. Secondly, based on objective EDR data analysis, it can provide a more scientific and intuitive basis for the determination of accident responsibility, promote the fairness and scientificity of traffic accident handling, reduce disputes among all parties, and protect the legitimate rights and interests of the parties.

[0033] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0035] Figure 1 is a flow chart of a traffic accident reconstruction method based on an automobile event data recording system according to an embodiment of the present invention;

[0036] Figure 2is a platform architecture diagram of a traffic accident reconstruction method based on an automobile event data recording system according to an embodiment of the present invention;

[0037] Figure 3 2 is a schematic diagram of the structure of a traffic accident re-device based on an automobile event data recording system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0038] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0039] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0040] The following describes a traffic accident reconstruction method and device based on an automobile event data recording system according to an embodiment of the present invention with reference to the accompanying drawings.

[0041] At present, in the research field of traffic accident scene investigation, there is still a lack of a comprehensive accident reconstruction platform based on EDR data. The present invention aims to combine EDR data with traditional data and verify each other through system design. According to the relevant provisions of the "Automobile Event Data Recording System", the present invention is based on traditional investigation data and uses EDR data as the technical core to carry out integrated accident reconstruction. The present invention particularly emphasizes the acquisition of EDR data, clarifies the scope of data acquisition, and adopts a more scientific method to efficiently and accurately obtain the working condition information of the vehicle during the incident, which will significantly improve the efficiency and accuracy of accident reconstruction. In terms of reconstruction ideas, the present invention proposes the concepts of "key completion" and "mutual verification". EDR information will serve as an important scientific supplement to traditional on-site investigation information; at the same time, on-site investigation data, such as vehicle damage, road surface marks, surveillance videos, etc., can all be used as the basis for verifying the "relevance, integrity, and validity" of EDR data. Information such as vehicle speed recorded in EDR data can also be used to verify the calculation of vehicle speed in traditional investigation or identification processes. The present invention systematically sorts out the process of accident reconstruction based on EDR and establishes a complete and scalable standardized accident reconstruction process. Through intelligent text recognition, fixed-point information capture, automatic docking model database and other functions, the automation level of accident reconstruction is greatly improved, thereby improving the efficiency of product technical work. The overall platform structure is as follows Figure 2 shown.

[0042] Figure 1 is a flow chart of a traffic accident reconstruction method based on an automobile event data recording system according to an embodiment of the present invention. Figure 1 As shown, including:

[0043] S1, obtaining on-site investigation data information, monitoring video image information and EDR data of the accident vehicle after the traffic accident scene.

[0044] It can be understood that the present invention obtains on-site investigation data information, including: static space information and accident-related information; monitoring video image information, including basic image information and information reflecting the accident; and the EDR data, including basic information and working condition information.

[0045] Specifically, the on-site investigation information after the traffic accident scene includes but is not limited to the road width, signs and markings, road facilities, vehicle damage, traces and scattered objects left on the road surface, damage to road facilities and buildings, injuries of the persons involved in the case, etc. Image information acquisition, fixed probe surveillance video and vehicle-mounted surveillance video information acquisition of the section where the accident occurred.

[0046] On-site investigation data information: static spatial information (road surface space dimensions, road signs and markings, road construction information, road surface adhesion coefficient); accident-related information (driver information, driving and passenger relationships in the vehicle, vehicle damage marks, and marks left on the road surface).

[0047] Monitoring video image information: basic image information (name and size, format and frame rate, integrity check value, whether it is clearly recorded); accident information (vehicle driving trajectory, vehicle collision process, braking before collision, special items before collision).

[0048] Specifically, the traffic accident information obtained includes:

[0049] Static spatial information: spatial dimensions of the road surface at the accident scene, the marking conditions and dimensions of road facilities and signs and markings; road slope, curvature, road adhesion coefficient and other information.

[0050] Information on damage to people, vehicles and roads after the accident: basic information of the driver, health status, drinking or drug use, driving and passenger relationship, etc.; basic vehicle information (including but not limited to model, size, weight, structure, center of gravity, material, etc.); damage to the vehicle after the accident and traces left on the road by tires and body, etc.; scattered objects on the road, bloodstains, damage to road facilities and buildings, etc.

[0051] Image information: image name, size, format, integrity check value, frame rate, pixels, etc.; the image records the vehicle’s driving trajectory, acceleration, deceleration and braking conditions, collision process, traces left at the accident scene after the collision, and other relevant information.

[0052] The vehicle operating condition information reflected by EDR includes: the vehicle's speed, braking, acceleration, steering angle, engine / motor speed within 5 seconds before the accident; the changes in the vehicle's lateral, longitudinal, and vertical acceleration during the accident; the status of the vehicle's assisted driving system (which may include cruise control, adaptive cruise, lane keeping, departure warning, lane change assist, collision warning, etc.).

[0053] EDR reflects the vehicle's passive and active safety system information: the working conditions of seat belts, airbags, etc.; automatic emergency braking system, body stability control system, anti-lock braking system, etc.

[0054] It is understandable that the full name of the vehicle event data recording system (EDR) is "Event Data Recorder". It is a system used to record data before and after a vehicle accident and is often referred to as the "black box" on the car. The main function of the EDR system is to record data when a vehicle collision or emergency occurs, including data before, during, and after the vehicle collision. These data include but are not limited to the vehicle's driving speed, steering wheel steering angle, engine operating status, accelerator pedal position, seat belt usage status, airbag status, driver assistance system status, and vehicle speed changes. EDR information will play an important role in accident investigation, accident cause analysis, accident prevention, and vehicle safety performance improvement.

[0055] The present invention uses a special EDR (Event Data Recorder) reading device to obtain EDR data, which is a key step in the process of traffic accident reconstruction. EDR is a vehicle-mounted device used to record key operating data and vehicle status information before and after a specific event (such as a collision). The present invention uses a special EDR reading device on the vehicle side or disassembles the device carrying EDR information to a laboratory to read and save data.

[0056] S2, data fusion and completion are performed on the on-site investigation data information, the monitoring video image information and the verified EDR data of the accident vehicle to obtain the data fusion result.

[0057] Specifically, the data fusion content involved in the present invention is to complete and filter data information to make the data information of the case more comprehensive. For example: on-site investigation records can obtain information such as road marks, vehicle collision sites, vehicle driving trajectories, etc., but cannot reflect information such as vehicle driving speed and the specific time of the accident. By completing the monitoring video content, the time of the incident can be determined and the vehicle driving speed can be obtained by calculation; but at the same time, whether the vehicle driving speed calculated based on the video is accurate, the braking or acceleration of the vehicle at the time of the accident, etc., cannot be determined; by continuing to fuse EDR information, the case information can be further completed and improved. However, whether the EDR information is valid and accurate requires data verification.

[0058] Among them, the relevance, integrity and validity of the EDR information are judged and verified: the present invention comprehensively considers the existing information of the accident to judge whether the EDR information is left over from this incident, whether the recorded data can fully reflect the content of the accident process, and whether the recorded data information (such as vehicle speed, direction angle, etc.) is objective and accurate. After further judgment, the judgment result can be used to determine whether the acquired EDR data can be used for on-site reconstruction.

[0059] In an embodiment of the present invention, the relevance, integrity, and validity of EDR information are determined and verified:

[0060] Relevance Verification:

[0061] Vehicle identification code consistency: The vehicle identification code recorded in the EDR data is consistent with the accident vehicle identification code, indicating that the EDR data comes from the accident vehicle.

[0062] Power-on cycle consistency: The power-on cycle at the time of the accident recorded in the EDR data is similar to the power-on cycle when the EDR data is read, which further confirms that the EDR data is related to the accident.

[0063] Consistency of airbag deployment: The airbag deployment recorded by the EDR data is consistent with the airbag detonation period shown in the accident vehicle inspection photos and on-board video, further supporting the correlation between the EDR data and this accident.

[0064] The collision shape and time interval are consistent: The collision shape and time interval of the two collisions recorded in the EDR data are consistent with those shown in the accident vehicle inspection photos, on-board videos, and on-site photos, indicating that the EDR data reflects the circumstances of the two collisions in this accident.

[0065] Integrity Verification:

[0066] Data record integrity: The event data recorded by the EDR data should all be shown as successfully and completely recorded, indicating that the EDR data fully reflects the movement conditions of the accident vehicle during the incident.

[0067] Validity Verification:

[0068] Tire specifications: The tire specifications of the accident vehicle at the time of the accident were in line with the factory settings, and there was no abnormal damage to the tires, eliminating the impact of tire specifications on the accuracy of EDR vehicle speed.

[0069] Collision analysis: The two collisions recorded by the EDR data are consistent with the accident vehicle inspection photos, on-board videos, and on-site photos, indicating that the EDR data effectively reflects the accident process.

[0070] Time node analysis: If the EDR data records two accident data, the time interval between the two collisions, the driving status in the vehicle, whether braking measures were taken, etc. can all be used to prove the validity of the EDR data.

[0071] In an embodiment of the present invention, EDR information is used to cross-validate the accuracy of traditional information.

[0072] Time consistency: Confirm whether the EDR recording time is consistent with the investigation record and surveillance video time.

[0073] Collision process: Compare the collision form, time and investigation traces recorded by EDR, and whether the surveillance video content is consistent.

[0074] Vehicle speed: Compare the EDR vehicle speed with the speed estimated by the survey and the speed estimated by the surveillance video to see if they are consistent.

[0075] Driver operation: Compare the driver operation data recorded by the EDR with the driver behavior in the investigation records and surveillance videos to see if they are consistent.

[0076] S3, based on the data fusion results, automatic modeling is performed to build a standardized accident reconstruction process to obtain the accident reconstruction results, and the accident reconstruction results are verified.

[0077] The present invention targets the above reorganized information and carries out reconstruction application of the above information in accordance with the definition of on-site reconstruction in the "Technical Specifications for Image-Based Road Traffic Accident Reconstruction" (SF / T 0160-2023).

[0078] The present invention mainly uses software such as PC-crash or COMSOL Multiphysics to construct and input static spatial information, and uses the damage traces of people, vehicles and roads added after the accident as the reconstruction target to reconstruct the accident process more accurately and scientifically.

[0079] It is understandable that intelligent text recognition: is applied to the extraction of text information in accident reports, maintenance logs and other documents. Through OCR technology (Optical Character Recognition), the text in paper documents or scanned copies can be converted into editable text format, which facilitates further data processing and analysis. Automatically extract key information, such as the time, location, and participants of the accident, reducing the workload and error rate of manual entry.

[0080] It can be understood that fixed-point information capture: by setting specific data capture rules, specific information related to the accident can be automatically extracted from sources such as databases and sensor records. For example, data records from a period of time before and after the accident can be extracted from the sensor network. It supports unified processing of data sources in different formats to ensure that all relevant information is fully captured.

[0081] It is understandable that the model library is automatically connected: a library containing various types of accident analysis models is established, and the most suitable analysis model is automatically selected according to the characteristics of different accidents. For example, for rock burst events, a prediction model designed specifically for such events can be selected. When new data enters the system, the appropriate model is automatically called for processing without manual specification or adjustment of the model configuration.

[0082] It is understandable that efficient automated modeling technology: using machine learning algorithms to automatically create or optimize models for accident analysis. This includes steps such as automatically selecting features, adjusting hyperparameters, and training models. Accelerating the model development process through automation technology allows the model to adapt to new data and scenarios more quickly, improving prediction accuracy and processing speed.

[0083] It is understandable that the standardized accident reconstruction process: Integrate the above technologies into a standardized accident reconstruction process to ensure that each accident analysis is carried out according to the same steps, thereby ensuring the consistency and comparability of the results. The process includes but is not limited to data collection, information extraction, model selection, analysis execution, and result interpretation, and each link has clear operating guidelines and technical support.

[0084] The application of these technologies helps to build an efficient accident handling system, which can not only quickly and accurately reconstruct the process of the accident, but also provide strong data support for future preventive measures. In this way, various potential safety hazards can be better understood and dealt with, reducing the probability and impact of accidents.

[0085] Specifically, the reconstruction step of the present invention specifically includes:

[0086] Data collection steps:

[0087] Text information processing: Use OCR technology and natural language processing technology to automatically identify and extract text information, and perform structured processing.

[0088] Automatic EDR data reading: Integrate with EDR data reading equipment to realize automatic reading and parsing of EDR data, and automatically extract important information related to modeling, such as vehicle speed, steering, braking, etc.

[0089] Data processing steps:

[0090] Image information processing: Use computer vision technology to pre-process surveillance video images and extract information such as vehicle trajectory and collision process.

[0091] EDR data processing: Parse EDR data and convert it into a format recognizable by PC-crash.

[0092] The preprocessed data are integrated and summarized to form a complete accident data set, and the data consistency is verified.

[0093] PC-crash modeling steps:

[0094] Model library management: Establish vehicle model library, scene model library, etc., and dynamically update and manage them.

[0095] Automatic modeling engine: According to the accident data set, it automatically matches the corresponding vehicle model and scene model, and sets parameters such as vehicle size, weight, material, collision speed, collision angle, etc.

[0096] Simulation engine: Run PC-crash simulation calculations, generate accident reconstruction animations, and analyze results.

[0097] Result analysis steps:

[0098] Analyze the accident reconstruction animation, obtain relevant parameters of the accident process, and display them visually.

[0099] Compare the accident reconstruction results with the investigation traces, surveillance video and EDR data to verify the accuracy of the reconstruction results.

[0100] Automatically generate accident reconstruction reports, including accident process description, parameter analysis, result conclusions, etc.

[0101] Furthermore, the reconstructed accident process needs to accurately apply the vehicle operating conditions before the accident reflected by the EDR and the impact of the passive safety system in the vehicle on the injury process of the occupants during the accident. The reconstructed results can be further compared and verified with the information reflected by the image information of the accident process to ensure the accuracy of the reconstruction results.

[0102] The application of this platform can further analyze the causes of accidents based on the accident process after the reconstruction results are given, and can make suggestions on possible causes of accidents, such as driver failure to drive in accordance with the law, driver distraction or fatigue, problems with vehicle braking or steering and other safety technical conditions, problems with road slope or curvature, problems with road adhesion coefficient or drainage function, etc. This will provide support for the improvement of accident rules and traffic regulations and the improvement of traffic facilities, thereby improving the level of road traffic safety and reducing the occurrence of traffic accidents.

[0103] According to the traffic accident reconstruction method based on the automobile event data recording system of the embodiment of the present invention, a complete and scalable standardized accident reconstruction process has been established. Through functions such as intelligent text recognition, fixed-point information capture, and automatic docking with model databases, the automation level of accident reconstruction has been greatly improved, thereby improving the efficiency of product technical work.

[0104] In order to implement the above embodiment, Figure 3 As shown, this embodiment also provides a traffic accident reconstruction device 10 based on the automobile event data recording system, including:

[0105] The data acquisition module 100 is used to acquire on-site investigation data information, monitoring video image information and EDR data of the accident vehicle after the traffic accident scene;

[0106] The data fusion module 200 is used to perform data fusion and completion on the on-site investigation data information, the monitoring video image information and the verified EDR data of the accident vehicle to obtain a data fusion result;

[0107] The accident reconstruction module 300 is used to perform automatic modeling based on the data fusion result to construct a standardized accident reconstruction process to obtain an accident reconstruction result, and verify the accident reconstruction result.

[0108] In one embodiment of the present invention, the on-site investigation data information includes: static space information and accident-related information; the monitoring video image information includes basic image information and information reflecting the accident; and the EDR data includes basic information and working condition information.

[0109] In one embodiment of the present invention, the static spatial information includes road surface spatial dimensions, road signs and markings, road construction information, and road surface adhesion coefficient;

[0110] The accident-related information includes driver information, driving and passenger relationships in the vehicle, signs of vehicle damage, and marks left on the road surface;

[0111] The basic information of the image, including name and size, format and frame rate, integrity check value, and whether it is recorded clearly;

[0112] The accident information includes the vehicle's driving trajectory, the vehicle's collision process, braking before the collision, and steering before the collision;

[0113] The basic information includes basic vehicle information, system status during reading, software and hardware used for reading, data record completeness, data source description and data limitations;

[0114] The operating condition information includes driving speed, longitudinal acceleration change, lateral acceleration change, collision type, vehicle braking condition, vehicle steering condition, vehicle acceleration condition, engine speed, seat belt and airbag working conditions, ABS and AEB working conditions, and ACC and LDW working conditions.

[0115] In one embodiment of the present invention, the EDR data of the accident vehicle is verified, including: relevance verification, integrity verification and validity verification; wherein,

[0116] The correlation verification is as follows: verification of the consistency between the vehicle identification code recorded in the EDR data and the accident vehicle identification code, so as to obtain the consistency result of the vehicle identification code; verification that the power-on cycle at the time of the accident recorded in the EDR data is similar to the power-on cycle when the EDR data is read, so as to obtain the consistency verification result of the power-on cycle; verification that the airbag deployment recorded in the EDR data is consistent with the airbag detonation period shown in the accident vehicle inspection photos and the on-board video, so as to obtain the consistency verification result of the airbag deployment; verification that the two collision forms and time intervals recorded in the EDR data are consistent with those shown in the accident vehicle inspection photos, on-board videos, and on-site photos, so as to obtain the collision form and time interval consistency verification result;

[0117] The integrity verification: verifying that the event data recorded in the EDR data should all be represented as successfully and completely recorded, so as to obtain the data record integrity verification result;

[0118] The validity verification: performs tire specification verification, collision analysis and time node analysis to obtain validity verification results.

[0119] In one embodiment of the present invention, the accident reconstruction module is further used to:

[0120] Use OCR technology and natural language processing technology to process on-site investigation data information to automatically identify and extract text information in a targeted manner, and perform structured processing; use computer vision technology to process surveillance video image information, and parse EDR data to convert it into a format recognizable by the PC-crash model to obtain pre-processed data; fuse the pre-processed data to obtain a complete accident data set;

[0121] Establish vehicle model library and scene model library, and dynamically update and manage them; automatically match the corresponding vehicle model and scene model according to the accident data set, and set parameters; run PC-crash simulation calculations to generate accident reconstruction animations;

[0122] Analyze the accident reconstruction animation, obtain relevant parameters of the accident process, and visualize the accident reconstruction results; compare the accident reconstruction results with investigation traces, surveillance videos and EDR data to verify the accuracy of the accident reconstruction results; automatically generate an accident reconstruction report.

[0123] According to the traffic accident reconstruction device based on the automobile event data recording system of the embodiment of the present invention, a complete and scalable standardized accident reconstruction process has been established. Through functions such as intelligent text recognition, fixed-point information capture, and automatic docking with model databases, the automation level of accident reconstruction has been greatly improved, thereby improving the efficiency of product technical work.

[0124] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0125] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

Claims

1. A traffic accident reconstruction method based on an automobile event data recording system, characterized in that: include: Obtain on-site investigation data information, surveillance video image information and EDR data of the accident vehicle after the traffic accident scene; Performing data fusion and completion on the on-site investigation data information, the monitoring video image information, and the verified EDR data of the accident vehicle to obtain a data fusion result; Based on the data fusion results, automated modeling is performed to construct a standardized accident reconstruction process to obtain accident reconstruction results, and the accident reconstruction results are verified.

2. The method according to claim 1, characterized in that The on-site investigation data information includes: static space information and accident-related information; the monitoring video image information includes basic image information and information reflecting the accident; the EDR data includes basic information and working condition information.

3. The method according to claim 2, characterized in that The static spatial information includes road surface spatial dimensions, road signs and markings, road construction information, and road surface adhesion coefficient; The accident-related information includes driver information, driving and passenger relationships in the vehicle, signs of vehicle damage, and marks left on the road surface; The basic information of the image, including name and size, format and frame rate, integrity check value, and whether it is recorded clearly; The accident information includes the vehicle's driving trajectory, the vehicle's collision process, braking before the collision, and steering before the collision; The basic information includes basic vehicle information, system status during reading, software and hardware used for reading, data record completeness, data source description and data limitations; The operating condition information includes driving speed, longitudinal acceleration change, lateral acceleration change, collision type, vehicle braking condition, vehicle steering condition, vehicle acceleration condition, engine speed, seat belt and airbag working conditions, ABS and AEB working conditions, and ACC and LDW working conditions.

4. The method according to claim 2, characterized in that: Verify the EDR data of the accident vehicle, including: relevance verification, integrity verification and validity verification; among them, The correlation verification is as follows: verification of the consistency between the vehicle identification code recorded in the EDR data and the accident vehicle identification code, so as to obtain the consistency result of the vehicle identification code; verification that the power-on cycle at the time of the accident recorded in the EDR data is similar to the power-on cycle when the EDR data is read, so as to obtain the consistency verification result of the power-on cycle; verification that the airbag deployment recorded in the EDR data is consistent with the airbag detonation period shown in the accident vehicle inspection photos and the on-board video, so as to obtain the consistency verification result of the airbag deployment; verification that the two collision forms and time intervals recorded in the EDR data are consistent with those shown in the accident vehicle inspection photos, on-board videos, and on-site photos, so as to obtain the collision form and time interval consistency verification result; The integrity verification: verifying that the event data recorded in the EDR data should all be represented as successfully and completely recorded, so as to obtain the data record integrity verification result; The validity verification: performs tire specification verification, collision analysis and time node analysis to obtain validity verification results.

5. The method according to claim 1, characterized in that: Based on the data fusion results, automatic modeling is performed to construct a standardized accident reconstruction process to obtain accident reconstruction results, including: Use OCR technology and natural language processing technology to process on-site investigation data information to automatically identify and extract text information in a targeted manner, and perform structured processing; use computer vision technology to process surveillance video image information, and parse EDR data to convert it into a format recognizable by the PC-crash model to obtain pre-processed data; fuse the pre-processed data to obtain a complete accident data set; Establish vehicle model library and scene model library, and dynamically update and manage them; automatically match the corresponding vehicle model and scene model according to the accident data set, and set parameters; run PC-crash simulation calculations to generate accident reconstruction animations; Analyze the accident reconstruction animation, obtain relevant parameters of the accident process, and visualize the accident reconstruction results; compare the accident reconstruction results with investigation traces, surveillance videos and EDR data to verify the accuracy of the accident reconstruction results; automatically generate an accident reconstruction report.

6. A traffic accident reconstruction device based on an automobile event data recording system, characterized in that: include: The data acquisition module is used to obtain on-site investigation data information, monitoring video image information and EDR data of the accident vehicle after the traffic accident scene; A data fusion module is used to perform data fusion and completion on the on-site investigation data information, the monitoring video image information and the verified EDR data of the accident vehicle to obtain a data fusion result; The accident reconstruction module is used to automatically model the accident based on the data fusion result to construct a standardized accident reconstruction process to obtain the accident reconstruction result, and verify the accident reconstruction result.

7. The device according to claim 6, characterized in that The on-site investigation data information includes: static space information and accident-related information; the monitoring video image information includes basic image information and information reflecting the accident; the EDR data includes basic information and working condition information.

8. The device according to claim 6, characterized in that The static spatial information includes road surface spatial dimensions, road signs and markings, road construction information, and road surface adhesion coefficient; The accident-related information includes driver information, driving and passenger relationships in the vehicle, signs of vehicle damage, and marks left on the road surface; The basic information of the image, including name and size, format and frame rate, integrity check value, and whether it is recorded clearly; The accident information includes the vehicle's driving trajectory, the vehicle's collision process, braking before the collision, and steering before the collision; The basic information includes basic vehicle information, system status during reading, software and hardware used for reading, data record completeness, data source description and data limitations; The operating condition information includes driving speed, longitudinal acceleration change, lateral acceleration change, collision type, vehicle braking condition, vehicle steering condition, vehicle acceleration condition, engine speed, seat belt and airbag working conditions, ABS and AEB working conditions, and ACC and LDW working conditions.

9. The device according to claim 7, characterized in that Verify the EDR data of the accident vehicle, including: relevance verification, integrity verification and validity verification; among them, The correlation verification is as follows: verification of the consistency between the vehicle identification code recorded in the EDR data and the accident vehicle identification code, so as to obtain the consistency result of the vehicle identification code; verification that the power-on cycle at the time of the accident recorded in the EDR data is similar to the power-on cycle when the EDR data is read, so as to obtain the consistency verification result of the power-on cycle; verification that the airbag deployment recorded in the EDR data is consistent with the airbag detonation period shown in the accident vehicle inspection photos and the on-board video, so as to obtain the consistency verification result of the airbag deployment; verification that the two collision forms and time intervals recorded in the EDR data are consistent with those shown in the accident vehicle inspection photos, on-board videos, and on-site photos, so as to obtain the collision form and time interval consistency verification result; The integrity verification: verifying that the event data recorded in the EDR data should all be represented as successfully and completely recorded, so as to obtain the data record integrity verification result; The validity verification: performs tire specification verification, collision analysis and time node analysis to obtain validity verification results.

10. The device according to claim 6, characterized in that The accident reconstruction module is further used for: Use OCR technology and natural language processing technology to process on-site survey data information to automatically identify and extract text information in a targeted manner, and perform structured processing; Use computer vision technology to process surveillance video image information and parse EDR data to convert it into a format recognizable by the PC-crash model to obtain pre-processed data; The preprocessed data is fused to obtain a complete accident data set; Establish vehicle model library and scene model library, and dynamically update and manage them; According to the accident data set, the corresponding vehicle model and scene model are automatically matched and parameter settings are performed; PC-crash simulation calculations are run to generate accident reconstruction animations; Analyze the accident reconstruction animation, obtain relevant parameters of the accident process, and visualize the accident reconstruction results; compare the accident reconstruction results with investigation traces, surveillance videos and EDR data to verify the accuracy of the accident reconstruction results; automatically generate an accident reconstruction report.

Images