Vehicle-mounted traffic accident rapid determination system

Abstract

The application discloses a vehicle-mounted traffic accident rapid judgment system and relates to the technical field of traffic management.A construction module is used for determining vehicle information, collecting driving environment information of the vehicle in real time based on sensors, and constructing an environment model in real time based on the driving environment information.A judgment module is connected with the construction module and is used for judging whether an accident occurs based on the environment model through machine learning and obtaining accident information when the accident occurs.A judgment module is connected with the judgment module and is used for completing accident judgment according to the accident information.The application can better construct and update the environment information in real time, and is convenient for analyzing and judging the vehicle accident subsequently.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of traffic management, in particular to a vehicle-mounted traffic accident rapid judgment system. BACKGROUND

[0002] With the rapid development of science and technology, various new technologies have been widely used in vehicle accident judgment, such as intelligent traffic system, vehicle-mounted sensor, image recognition technology, etc. However, the construction of high-precision virtual environment model requires rapid processing and analysis of a large amount of complex data. The existing computing platform and algorithm often face performance bottleneck when processing large-scale, high-dimensional data, resulting in a long time-consuming in the model construction process, which is difficult to meet the real-time requirements. At the same time, the accuracy of model construction is restricted by multiple factors such as data quality and algorithm complexity, which is difficult to maintain stable performance in complex road environment, and difficult to make rapid judgment of vehicle accidents according to vehicle driving conditions. SUMMARY

[0003] The purpose of the present application is to provide a vehicle-mounted traffic accident rapid judgment system to solve the problems in the background art.

[0004] In order to achieve the above purpose, the present application provides the following technical scheme: a vehicle-mounted traffic accident rapid judgment system, comprising:

[0005] A construction module is used to determine vehicle information, real-time collect driving environment information of the vehicle based on sensors, and real-time construct an environment model based on the driving environment information;

[0006] A judgment module is connected with the construction module, and is used to judge whether an accident occurs based on the environment model by machine learning, and obtain accident information when an accident occurs;

[0007] A judgment module is connected with the construction module, and is used to judge whether an accident occurs based on the environment model by machine learning, and obtain accident information when an accident occurs;

[0008] In one preferred embodiment, the construction module comprises:

[0009] An acquisition unit is used to determine vehicle type information and vehicle identity information as vehicle information;

[0010] A driving information acquisition unit is used to real-time collect driving environment information of the vehicle through the in-vehicle sensor, wherein the driving environment information includes vehicle surrounding environment information and vehicle driving parameter information;

[0011] A model construction unit is used to real-time construct an environment model by information fusion of the driving environment information, and bind the environment model with the corresponding vehicle information.

[0012] In a preferred implementation, the model construction unit comprises:

[0013] An information determination unit is configured to determine a driving jurisdiction, collect fixed environment information in the driving jurisdiction, and construct a fixed model based on the fixed environment information.

[0014] An update unit is configured to mark the fixed model to obtain fixed model points, and update the fixed model based on the fixed model points and the vehicle surrounding environment information to obtain an updated model.

[0015] A fusion unit is configured to construct a vehicle model based on the vehicle self-driving parameter information, construct the vehicle model in the updated model to complete the fusion of the driving environment information to obtain an environment model, and bind the environment model with corresponding vehicle information.

[0016] In a preferred implementation, the update unit comprises:

[0017] A division unit is configured to divide the fixed model to obtain a plurality of management ranges, mark points in the plurality of management ranges, configure cloud servers in the plurality of management ranges, and connect the cloud servers in the plurality of management ranges.

[0018] A setting unit is configured to obtain the number of environment subjects in the plurality of management ranges and images corresponding to the environment subjects, set update points in the environment subjects in the fixed model, and connect the update points with the cloud servers corresponding to the range points in the management ranges.

[0019] A management unit is configured to manage the management ranges connected by the cloud servers corresponding to the update points and the range points in the fixed model as fixed model points.

[0020] A model update unit is configured to extract images in the vehicle surrounding environment information to obtain images corresponding to a plurality of current environment subjects, compare the images corresponding to the plurality of current environment subjects with images corresponding to environment subjects in the corresponding fixed model to obtain comparison results, update the fixed model based on images corresponding to current environment subjects that do not meet preset conditions to obtain an updated model.

[0021] In a preferred implementation, the setting unit comprises:

[0022] A range division unit is configured to extract edge lines of the images corresponding to the environment subjects, limit image frames of the images corresponding to the environment subjects based on the edge lines, divide the images corresponding to the environment subjects to obtain a plurality of color blocks, and configure a cloud server for each color block.

[0023] An image configuration unit is configured to configure a plurality of color blocks of a single environmental subject and images corresponding to the color blocks as a single update point in a cloud server.

[0024] A connection unit is configured to connect the update point and the cloud server corresponding to the range point in the management range.

[0025] In a preferred embodiment, the judgment module comprises:

[0026] A training unit is configured to collect historical accident information, wherein the historical accident information comprises historical driving environment information and historical accident information, and the trained learning model is obtained by training the deep learning model based on the historical accident information.

[0027] A determination unit is configured to bind the trained learning model and the environmental model, input the driving environment information in the environmental model into the trained learning model, and output the accident information of the current vehicle.

[0028] In a preferred embodiment, the determination module comprises:

[0029] A process setting unit is configured to set a claim settlement process library, and select a corresponding claim settlement process according to the accident information of the current vehicle.

[0030] A determination application unit is configured to determine the accident information according to the claim settlement process, and store the accident information and the vehicle information corresponding to the claim settlement process.

[0031] In a preferred embodiment, the process setting unit comprises:

[0032] A database setting unit is configured to set the claim settlement process and the corresponding accident information as the claim settlement process library, wherein the claim settlement process comprises a responsible party claim settlement process and a non-responsible party claim settlement process.

[0033] A selection unit is configured to select a corresponding claim settlement process according to the accident information of the current vehicle.

[0034] In the above technical solution, the present application provides technical effects and advantages:

[0035] The present application can better quickly construct and update the environmental information by the cloud server, and facilitate the subsequent analysis and determination of the vehicle accident. BRIEF DESCRIPTION OF DRAWINGS

[0036] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments. Obviously, the drawings described below are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art based on these drawings.

[0037] Figure 1 The system block diagram of the present application. DETAILED DESCRIPTION

[0038] In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the following will combine the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the present application.

[0039] Embodiment 1, please refer to Figure 1 The vehicle-mounted traffic accident rapid judgment system described in this embodiment comprises:

[0040] The construction module is used to determine the vehicle information, collect the driving environment information of the vehicle in real time based on the sensor, and construct the environment model in real time based on the driving environment information;

[0041] The sensor is a device for interaction between the vehicle and the external environment, which can capture and transmit multi-dimensional information including road conditions, traffic flow, weather conditions, other road user dynamics, etc. By integrating multiple sensors (such as high-definition cameras, radars, laser radars, ultrasonic sensors, etc.), the vehicle can construct a three-dimensional and comprehensive driving environment perception network. The collected driving environment information of the vehicle and the vehicle can be interconnected, and the collected information can be used together to construct the environment model;

[0042] The judgment module is connected with the construction module, and is used to judge whether an accident occurs based on machine learning on the environment model, and obtain accident information when an accident occurs;

[0043] After the vehicle driving environment model is constructed, the machine learning algorithm is introduced to deeply analyze the massive data in these models to intelligently judge whether an accident occurs. This process not only improves the efficiency of accident judgment, but also significantly enhances the accuracy and objectivity of judgment;

[0044] The judgment module is connected with the construction module, and is used to judge whether an accident occurs based on machine learning on the environment model, and obtain accident information when an accident occurs;

[0045] The machine learning system accurately extracts accident information from the environment model, which becomes important information for starting the claim process. This information not only records the specific details of the accident, but also provides direct evidence for subsequent liability division, loss assessment, and compensation scheme formulation. The first step of the claim process is to verify and organize the accident information. The claim officer will carefully review the accident information, including the time, location, and circumstances of the parties involved, the extent of vehicle damage, and the status of personnel casualties, to ensure the completeness and accuracy of the information. This process may also involve communication with the parties involved, witnesses, and on-site investigators to obtain more supplementary evidence. Subsequently, the liability division stage is entered. Based on the accident information, traffic rules, and laws and regulations, the claim officer will comprehensively assess the responsibilities of each party and issue a liability determination. This step is the core of the claim process and directly affects the allocation of compensation responsibilities and the determination of compensation amounts. After the liability is clear, the loss assessment and compensation calculation stage is entered. The claim officer will quantify the losses caused by the accident based on vehicle repair quotes, medical expense lists, and property loss certificates, and calculate a reasonable compensation amount accordingly. Finally, the compensation scheme is formulated and implemented. The claim officer will formulate a detailed compensation scheme based on the loss assessment results and the insurance company's compensation policy, and negotiate with the parties involved to confirm the scheme. Once the compensation scheme is recognized, the insurance company will quickly start the compensation process to ensure that the victims can receive the compensation they deserve;

[0046] It should be noted that with the rapid development of technology, various new technologies have been widely applied in vehicle accident determination; for example, intelligent transportation systems, vehicle-mounted sensors, image recognition technology, etc. At present, in order to build a high-precision virtual environment model, a large amount of complex data needs to be processed and analyzed quickly. The existing computing platforms and algorithms often face performance bottlenecks when processing large-scale, high-dimensional data, resulting in a long time-consuming model construction process that cannot meet real-time requirements. At the same time, the accuracy of model construction is restricted by multiple factors such as data quality and algorithm complexity, making it difficult to maintain stable performance in complex and variable road environments, and making it difficult to quickly determine vehicle accidents based on vehicle driving conditions. In the present application, the cloud server sends the difference information of the comparison result to the image configuration cloud at the corresponding position, changes the corresponding color block through the image configuration cloud, updates the color block, and then obtains the updated fixed model, which can better quickly construct and update the environmental information, facilitating subsequent analysis and determination of vehicle accidents;

[0047] In one embodiment, the construction module comprises:

[0048] The acquisition unit is configured to determine vehicle type information and vehicle identity information as vehicle information;

[0049] The driving information acquisition unit is configured to acquire real-time driving environment information of the vehicle through a sensor built in the vehicle, wherein the driving environment information comprises vehicle surrounding environment information and vehicle self-driving parameter information.

[0050] The model construction unit is configured to fuse the driving environment information to construct a real-time environment model, and bind the environment model with corresponding vehicle information.

[0051] It should be noted that each vehicle has its own vehicle information, which comprises vehicle type information and vehicle identity information. The vehicle type information is mechanical information of the vehicle itself, such as vehicle brand and model. After determining the vehicle information, the vehicle surrounding environment information and the vehicle self-driving parameter information are acquired in real time through the sensor built in the vehicle during the driving of the vehicle as the driving environment information. The sensor built in the vehicle comprises a speed sensor and collection devices such as a radar, a laser radar and a camera. For example, when there is relative motion between the radar and a target, the frequency of the echo signal received by the radar will change. The signal processing system of the radar processes the received signal by FFT (Fast Fourier Transform) to extract the Doppler frequency and calculate the speed of the target. Then, the surrounding environment information of the vehicle is collected by the collection devices, such as the road environment where the vehicle is located, the facilities around the road, and other vehicles and pedestrians around the vehicle. The self-driving parameter information of the vehicle during the driving, such as the driving speed, direction and braking state of the vehicle, is also collected. Thus, multiple information related to the vehicle and the surrounding environment during the driving of the vehicle can be obtained. Then, the driving environment information is fused to construct a real-time environment model. Then, the environment model is bound with corresponding vehicle information.

[0052] In one embodiment, the model construction unit comprises:

[0053] The information determination unit is configured to determine a driving jurisdiction range, acquire fixed environment information in the driving jurisdiction range, and construct a fixed model based on the fixed environment information.

[0054] The update unit is configured to mark the fixed model to obtain fixed model points, and update the fixed model based on the fixed model points to obtain an updated model.

[0055] The fusion unit is configured to construct a vehicle model according to the vehicle self-driving parameter information, construct the vehicle model in the updated model to complete the fusion of the driving environment information to obtain an environment model, and bind the environment model with corresponding vehicle information.

[0056] It should be noted that the geographical range for determining the traffic accident judgment is determined as the driving jurisdiction range, and the fixed environment information in the driving jurisdiction range is collected, wherein the fixed environment information includes road information, road facility information and surrounding building information, which is basically fixed and unchanged. The above is collected in the form of image, so as to be a fixed model. The fixed model points are obtained by marking the fixed model. In the process of vehicle driving, the vehicle surrounding environment information is updated based on the fixed model points to obtain an updated model, which can determine the position of the vehicle and perform local update, thereby accelerating the update efficiency. Then, a vehicle model is constructed according to the vehicle driving parameter information. The vehicle model is a virtual vehicle shape model. The specific operation is to use a three-dimensional modeling software (such as Blender, Maya, 3ds Max, etc.) to create the appearance and shape of the vehicle, including the vehicle body, wheels, lights, etc. The vehicle model is generated by defining the parameters (such as length, width, height, wheelbase, etc.) and rules of the vehicle. Then, the vehicle shape model is virtually driven according to the vehicle driving parameter information and the actual vehicle driving condition. The vehicle model is constructed in the updated model to complete the fusion of the driving environment information to obtain an environment model. The virtually driven vehicle model is put into the updated model to obtain an environment model virtually driven in a virtual environment. The driving environment information is well fused through the environment model, which can virtually demonstrate the actual vehicle driving. Then, the environment model and the corresponding vehicle information are bound, which can facilitate subsequent vehicle accident judgment and analysis.

[0057] In one embodiment, the updating unit comprises:

[0058] The dividing unit is configured to divide the fixed model to obtain a plurality of management ranges, mark points in the plurality of management ranges as range points, and configure the range points in the plurality of management ranges to cloud servers.

[0059] The setting unit is configured to obtain the number of environment subjects in the plurality of management ranges and the images corresponding to the environment subjects, set update points in the environment subjects in the fixed model, and connect the update points and the cloud servers corresponding to the range points in the management ranges.

[0060] The management unit is configured to manage the management ranges connected by the cloud servers corresponding to the update points and the range points in the fixed model as fixed model points.

[0061] The model updating unit is configured to extract images in the vehicle surrounding environment information to obtain images corresponding to a plurality of current environment subjects, compare the images corresponding to the plurality of current environment subjects with images corresponding to environment subjects in the corresponding fixed model to obtain comparison results, and update the fixed model with images corresponding to current environment subjects that do not meet preset conditions to obtain an updated model.

[0062] It should be noted that the model construction needs to be divided into blocks in the entire driving jurisdiction range. There are multiple vehicles driving in one road, and therefore, the fixed model is updated by the collection of the vehicles, which facilitates the use of the updated model by all vehicles passing through the road and the analysis and determination of subsequent accidents. The division of the driving jurisdiction range also needs to be divided into ranges in the fixed model. After the fixed model is divided, a plurality of management ranges of the fixed model are obtained. Then, a range point is marked in each of the plurality of management ranges. The range point is used to manage the fixed environment information in the management range. Then, a cloud server is configured corresponding to the range point. The cloud servers are then connected in communication, so that the plurality of management ranges can be interconnected to form a complete fixed model. Then, environment subject data and images corresponding to the environment subjects in the plurality of management ranges are obtained. The environment subject is a type of object in the environment, such as a road in the fixed environment information and different subject types such as devices around the road. Then, the images corresponding to the environment subjects can be obtained. Therefore, the environment subjects in the management range need to be managed by the update point for subsequent updates of the environment subjects, so as to complete the update of the fixed model. The connection between the update point and the cloud server corresponding to the range point in the management range can be used to receive the part of the model that needs to be updated issued by the subsequent cloud server. After the management range connected by the update point and the cloud server corresponding to the range point in the fixed model is set as a fixed model point, the images in the vehicle surrounding environment information actually collected by the vehicle can be extracted to obtain current environment subject images. Then, the images corresponding to the plurality of current environment subjects are compared with the images corresponding to the environment subjects in the corresponding fixed model to obtain comparison results. The images corresponding to the plurality of current environment subjects and the images corresponding to the environment subjects in the corresponding fixed model are subjected to feature extraction (such as edges, corner points, and textures) by using algorithms such as SIFT (Scale-Invariant Feature Transform) and SURF (Speeded-Up Robust Features), and feature matching is performed to evaluate the similarity of the images. Then, the fixed model is updated with the images corresponding to the current environment subjects that do not meet the preset conditions to obtain an updated model, which can better complete the update of the model. When an accident occurs, the updated model can be used to analyze the mode and characteristics of the accident, which facilitates the determination of the accident and makes the determination of the accident more accurate and fast.

[0063] In one embodiment, the setting unit comprises:

[0064] a range division unit configured to extract an edge line of an image corresponding to an environmental subject, define an image frame of the image corresponding to the environmental subject based on the edge line, divide the image corresponding to the environmental subject to obtain a plurality of color blocks, and correspond each color block to an image configuration cloud;

[0065] an image configuration unit configured to configure the plurality of color blocks of a single environmental subject and the image configuration cloud corresponding to the color blocks as a single update point;

[0066] a connection unit configured to connect the update point and a cloud server corresponding to a range point in a management range;

[0067] It should be noted that the edge line of the image corresponding to the environmental subject is used as the image frame, and the image corresponding to the environmental subject is divided into a plurality of color blocks under the image frame, and one color block is configured to correspond to one image configuration cloud. The image configuration cloud is an image parameter library. For example, in a subsequent operation, if a color block is a full white picture and needs to be updated to red, the image configuration cloud will update the corresponding color block from white to red. The plurality of color blocks of a single environmental subject and the image configuration cloud corresponding to the color blocks are configured as a single update point. There are a plurality of environmental subjects in a single management range of a fixed model, and therefore there are a plurality of update points in a management range. Then, the update point and a cloud server corresponding to a range point in a management range are connected to receive an image update of the update point corresponding environmental subject by a subsequent range point corresponding cloud server, and then the update of the subsequent fixed model is completed to obtain an update module. In the model update unit, the images in the vehicle surrounding environment information are extracted to obtain a plurality of images corresponding to current environmental subjects. The images corresponding to the current environmental subjects are compared with the images corresponding to the environmental subjects in the corresponding fixed model to obtain a comparison result. The comparison result includes difference information and an occupied range of the difference information in an actual environment. The image corresponding to the current environmental subject corresponding to the comparison result that does not meet a preset condition is used to update the fixed model. The preset condition includes a preset occupied range. The update method is as follows: the cloud server sends the difference information of the comparison result to the image configuration cloud at a corresponding position, the image configuration cloud changes the corresponding color block, the update of the color block is completed, and then the updated fixed model is obtained. The updated fixed model can better quickly construct and update the environmental information, and is convenient for subsequent analysis and determination of vehicle accidents.

[0068] In one embodiment, the judgment module comprises:

[0069] a training unit configured to collect historical accident information, wherein the historical accident information includes historical driving environment information and historical accident information, and the historical accident information is used to train a deep learning model to obtain a trained learning model;

[0070] The determination unit is configured to bind the trained learning model with the environment model, input driving environment information in the environment model into the trained learning model, and output accident information of the current vehicle;

[0071] It should be noted that the historical accident information of all vehicles is collected, the historical accident information includes historical driving environment information and historical accident information, the historical accident information includes historical accident type and historical accident responsibility division, the historical accident information is cleaned to remove noise data, repeated data and incomplete data, non-numeric data is converted into numerical data, such as converting text data into numerical coding, and finally the data is standardized or normalized to obtain processed historical accident information. The processed data and labeled information are input into the deep learning model for training, the hyperparameters (such as learning rate, batch size, iteration number, etc.) of the model are optimized to obtain a trained learning model. When the trained learning model is used, the trained learning model is bound with the environment model, the driving environment information in the environment model is input into the trained learning model, and the accident information of the current vehicle is output. The accident information is accident feature information, and the accident feature information here is accident type and accident responsibility division. The accident responsibility division is divided into responsible party and non-responsible party. The damage of the vehicle can also be collected to obtain the accident repair information including the estimated value and the maintenance period from the vehicle damage estimation database. The learning model and the environment model can be combined and bound to quickly determine the vehicle accident, and the efficiency of processing the vehicle accident is improved.

[0072] In one embodiment, the determination module comprises:

[0073] The flow setting unit is configured to set a claim process library, and select a corresponding claim process according to the accident information of the current vehicle.

[0074] The determination application unit is configured to apply for determination of the accident information according to the claim process, and store the accident information and the vehicle information corresponding to the claim process.

[0075] In one embodiment, the flow setting unit comprises:

[0076] The database setting unit is configured to set the claim process and the corresponding accident information as the claim process library, wherein the claim process includes the responsible party claim process and the non-responsible party claim process.

[0077] The selection unit is configured to select a corresponding claim process according to the accident information of the current vehicle.

[0078] It should be noted that this library is divided into two core parts: the liability party claim settlement process and the non-liability party claim settlement process, aiming to cover all possible accident scenarios under different liability attribution; the liability party claim settlement process focuses on guiding the vehicle owner or insurance company identified as the party responsible for the accident to submit to the insurance company for review, and then according to the insurance terms, the claim settlement calculation and negotiation are carried out, and finally the claim settlement procedure is completed; the non-liability party claim settlement process provides clear guidance for vehicle owners who are not involved in the accident or only bear secondary responsibility.

[0079] When a vehicle accident occurs, the system will quickly respond to intelligent matching and select the most suitable claim settlement process to ensure the accuracy and efficiency of the claim settlement process selection; the system will guide the user to apply and determine the accident information step by step according to the selected claim settlement process, and the system will automatically or with manual review, comprehensive evaluation of the authenticity of the accident, the extent of the loss and the attribution of responsibility; in order to ensure the transparency and traceability of the claim settlement process, all accident information and its corresponding claim settlement process records, as well as key information related to the vehicle (such as vehicle identification number, insurance policy number, vehicle owner information, etc.) will be safely stored in the system database, building a claim settlement process library containing liability and non-liability claim settlement processes, and through intelligent matching and guiding mechanism to realize the application and determination of accident information and storage, not only improves the efficiency and accuracy of claim settlement processing, but also provides customers with a more convenient and transparent claim settlement service experience.

[0080] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A vehicle-mounted rapid accident determination system, characterized in that, include: The module is used to determine vehicle information, collect vehicle driving environment information in real time based on sensors, and perform information fusion based on the driving environment information to build an environment model in real time. The judgment module, connected to the construction module, is used to judge whether an accident has occurred based on the environment model using machine learning, and to obtain accident information when an accident has occurred. The determination module, connected to the judgment module, is used to complete the accident determination based on the accident information and the claims process. The building module includes: The data collection unit is used to determine vehicle type information and vehicle identity information as vehicle information. The driving information acquisition unit is used to collect real-time driving environment information of the vehicle through the built-in sensors of the vehicle. The driving environment information includes information about the environment around the vehicle and the vehicle's own driving parameters. The model building unit is used to fuse driving environment information to build an environment model in real time and bind the environment model with the corresponding vehicle information. The model building unit includes: The information determination unit is used to determine the driving jurisdiction area, collect fixed environmental information within the driving jurisdiction area, and construct a fixed model based on the fixed environmental information; The update unit is used to mark points on the fixed model to obtain the fixed model points, and to update the fixed model with the surrounding environmental information of the vehicle based on the fixed model points to obtain the updated model. The fusion unit is used to construct a vehicle model based on the vehicle's own driving parameter information, integrate the vehicle model into the update model to complete the fusion of driving environment information to obtain an environment model, and bind the environment model with the corresponding vehicle information. The update unit includes: The division unit is used to divide the scope in a fixed model to obtain multiple management scopes. Marking points are set in each of the multiple management scopes as scope points. Cloud servers are configured for the scope points in the multiple management scopes. Communication connections are established between the cloud servers in the multiple management scopes. The setting unit is used to obtain the number of environmental subjects and the corresponding images of the environmental subjects within multiple management ranges. Update points are set in the environmental subjects in the fixed model, and the update points are connected to the cloud servers corresponding to the range points within the management range. The management unit is used to connect the update point and the range point in the fixed model to the cloud server and then use the management range as the fixed model point location. The model update unit is used to extract images from the vehicle's surrounding environment information to obtain multiple images corresponding to the current environmental subjects, compare the multiple images corresponding to the current environmental subjects with the images corresponding to the environmental subjects in the corresponding fixed model to obtain comparison results, and update the fixed model with the images corresponding to the current environmental subjects corresponding to the comparison results that do not meet the preset conditions to obtain the updated model. The setting unit includes: The range division unit is used to extract the edge lines of the image corresponding to the main subject of the environment, define the image frame of the image corresponding to the main subject of the environment based on the edge lines, divide the image corresponding to the main subject of the environment into multiple color blocks, and assign each color block to an image configuration cloud. An image configuration unit is used for multiple color blocks of a single environmental subject and the image configuration cloud corresponding to the color blocks as a single update point; The connection unit is used to connect the update point with the cloud server corresponding to the range point within the management scope; The judgment module includes: The training unit is used to collect historical accident information, which includes historical driving environment information and historical accident information. The deep learning model is trained using the historical accident information to obtain a trained learning model. The decision unit is used to bind the trained learning model with the environment model, input the driving environment information in the environment model into the trained learning model and output the current vehicle's accident information. The determination module includes: The process setting unit is used to set up the claims process library and select the corresponding claims process based on the current vehicle's accident information. The application unit is used to apply for and determine the accident information according to the claims process, and to store the accident information and the corresponding vehicle information in the claims process. The process setting unit includes: The database setting unit is used to set up the claims process and corresponding accident information as a claims process database. The claims process includes claims process for the responsible party and claims process for the non-responsible party. The selection unit is used to select the corresponding claims process based on the current vehicle's accident information.

Images