WO2020000834A1 - 交通事故处理方法、系统及服务器 - Google Patents
交通事故处理方法、系统及服务器 Download PDFInfo
- Publication number
- WO2020000834A1 WO2020000834A1 PCT/CN2018/113400 CN2018113400W WO2020000834A1 WO 2020000834 A1 WO2020000834 A1 WO 2020000834A1 CN 2018113400 W CN2018113400 W CN 2018113400W WO 2020000834 A1 WO2020000834 A1 WO 2020000834A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- traffic accident
- information
- scene
- rescue
- designated
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Ceased
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Definitions
- the present application relates to the field of communication technologies, and in particular, to a method, a system, and a server for processing a traffic accident.
- the handling process is as follows: the relevant personnel of the traffic accident report the incident. After receiving the report, the transportation department and the insurance company assign personnel to the scene of the traffic accident to conduct an on-site investigation (such as taking pictures). Liability, issue a certificate of responsibility for traffic accidents. Alternatively, the traffic police subsequently retrieved the surveillance video of the accident scene and combined the site investigation information to determine the responsibility of the accident a few days after the occurrence of the traffic accident, and issued a traffic accident responsibility certificate.
- the embodiments of the present application provide a method, a system, and a server for processing a traffic accident, which are used to solve the problems of high-cost and time-consuming labor in the process of handling a traffic accident in the prior art.
- an embodiment of the present application provides a method for processing a traffic accident, where the method includes:
- Obtaining a specified accident analysis model which is obtained based on a deep learning algorithm and trained by using known traffic accident scene information and known traffic accident responsibility identification information corresponding to the known traffic accident scene information;
- traffic accident responsibility determination information is generated according to the traffic accident scene information.
- an embodiment of the present application provides a traffic accident processing system, where the system includes:
- An acquisition module configured to acquire a designated accident analysis model, which is based on a deep learning algorithm and is obtained by training using known traffic accident scene information and known traffic accident responsibility identification information corresponding to the known traffic accident scene information. of;
- a generating module is configured to generate traffic accident responsibility determination information based on the specified accident analysis model and according to the traffic accident scene information.
- an embodiment of the present application provides a server, where the server includes:
- a memory for storing the processor-executable instructions
- the processor is configured to:
- Obtaining a specified accident analysis model which is obtained by training based on a deep learning algorithm using known traffic accident scene information and known traffic accident responsibility identification information corresponding to the known traffic accident scene information;
- traffic accident responsibility determination information is generated according to the traffic accident scene information.
- an embodiment of the present application provides a computer non-volatile readable storage medium.
- the computer non-volatile readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented:
- Obtaining a specified accident analysis model which is obtained by training based on a deep learning algorithm using known traffic accident scene information and known traffic accident responsibility identification information corresponding to the known traffic accident scene information;
- traffic accident responsibility determination information is generated according to the traffic accident scene information.
- the designated accident analysis model by receiving traffic accident scene information, obtaining a designated accident analysis model trained based on a deep learning algorithm, based on the designated accident analysis model, and generating traffic accident responsibility identification information based on the traffic accident scene information, the designated accident analysis model can be used. Automatically and intelligently generate traffic accident responsibility identification information, no traffic police and insurance personnel are required to visit the scene, and no traffic accident personnel are required to wait at the scene. On the one hand, it saves manpower costs, reduces the time for accident handling, and improves the speed of accident handling. On the one hand, the waiting time of traffic accident personnel is greatly reduced, so that traffic accident personnel without major disturbance can quickly leave the accident scene, avoiding causing traffic congestion or shortening the congestion time.
- FIG. 1 is a first example flowchart of a traffic accident processing method according to an embodiment of the present application.
- FIG. 2 is a second example flowchart of a traffic accident processing method provided by an embodiment of the present application.
- FIG. 3 is a third exemplary flowchart of a traffic accident processing method according to an embodiment of the present application.
- FIG. 4 is a fourth exemplary flowchart of a traffic accident processing method provided by an embodiment of the present application.
- FIG. 5 is a structural block diagram of a traffic accident processing system according to an embodiment of the present application.
- FIG. 6 is a simplified block diagram of a server provided by an embodiment of the present application.
- the embodiment of the present application provides a method for processing a traffic accident.
- the traffic accident processing method provided in the embodiment of the present application may be executed on a server of a transportation department.
- FIG. 1 is a first example flowchart of a traffic accident processing method according to an embodiment of the present application. As shown in FIG. 1, in this embodiment, the method for handling a traffic accident may include the following steps:
- S101 Receive traffic accident scene information.
- the specified accident analysis model is based on a deep learning algorithm and is obtained by training using known traffic accident responsibility information corresponding to known traffic accident scene information and known traffic accident scene information.
- the traffic accident responsibility identification information is generated according to the traffic accident scene information.
- the traffic accident scene information may include pictures and videos of the traffic accident scene.
- the accident scene environment, the scene photos of the vehicles involved, and monitoring video can be used as traffic accident scene information.
- the source of the traffic accident scene information received in step S101 may be a traffic monitoring device on the road section where the traffic accident scene is located, or a personal handheld device such as a mobile phone or a tablet of the car owner involved.
- the traffic monitoring equipment on the road segment where the traffic accident is located can collect the scene picture in real time, and upload the scene picture to the server of the transportation department together with the stored video of the time period of the traffic accident.
- the server can receive traffic accident scene information.
- the operation of collecting and uploading pictures and / or videos by the traffic monitoring device may be actively performed based on the judgment result of its own judgment program, or it may be passively triggered upon receiving an instruction issued by the background system of the traffic department.
- the method for receiving traffic accident scene information may be: receiving traffic accident scene information collected by a traffic monitoring device on a road section where the traffic accident scene is located.
- car owners also known as traffic accident parties
- passers-by, etc. can take live pictures or live videos through personal handheld devices such as mobile phones and tablets, and then send live pictures and / or live videos through the network Send to the server of the transportation department.
- the server may be provided with a dedicated interface for receiving information uploaded by personal handheld devices of the car owner and passerby involved.
- the interface may be a data upload interface of the official website of the transportation department, or an application (Application) installed on a personal handheld device, or an interface subroutine on the application.
- the method for receiving traffic accident scene information may be: receiving traffic accident scene information uploaded by a personal handheld device.
- the background server of the transportation department can quickly and easily obtain the information of the traffic accident scene without having to assign personnel to the scene of the traffic accident to conduct a manual on-site survey, which will help shorten the processing time of the accident and improve the efficiency of the accident. It also saves labor costs.
- the designated accident analysis model in step S102 may be pre-trained.
- obtaining a specified accident analysis model may include: obtaining multiple sets of training data, each set of training data including known traffic accident scene information and known traffic accidents corresponding to the known traffic accident scene information Responsibility determination information; extract known traffic accident scene information from multiple sets of training data as input data for a designated accident analysis model; extract known traffic accident responsibility determination information from multiple sets of training data as output data for a specified accident analysis model ; Use input data and output data to train a deep learning algorithm model to generate a designated accident analysis model.
- the acquisition process of the specified accident analysis model may be: 10,000 sets of known traffic accident scene information that has occurred and known traffic accidents corresponding to the known traffic accident scene information
- Responsibility identification information is used as training data.
- Several parameters are set for the convolutional neural network model. These parameters are unknown before training.
- the known traffic accident scene information in the training data is used as the input of the convolutional neural network model.
- the known traffic accident liability identification information is used as the output of the convolutional neural network model to solve the parameter values of the parameters. At this time, the parameter values of these parameters are known. These parameter values determine a specific volume.
- Product neural network model which can be used as a designated accident analysis model.
- step S102 generating traffic accident responsibility determination information based on the traffic accident scene information based on the designated accident analysis model may include: inputting traffic accident scene information into the designated accident analysis model as input data; obtaining output data of the designated accident analysis model , Use this output data as traffic accident liability identification information.
- the traffic accident liability confirmation information may be a traffic accident liability confirmation.
- the content of traffic accident liability identification information may include:
- the specific content contained in the traffic accident liability determination information can be adjusted according to needs or actual conditions, such as adding, reducing, or modifying the corresponding content.
- the background server of the transportation department can automatically and intelligently analyze and judge the traffic accident, and then determine the responsibility of the traffic accident. In this way, there is no longer a need for traffic police to rush to the scene of a traffic accident, and the responsibility for a traffic accident can be determined manually, thereby reducing the time for handling the accident and increasing the speed of handling the accident.
- steps S101 and S102 it can be known that the embodiment of the present application can automatically and intelligently blame a traffic accident based on remotely collected traffic accident scene information. Therefore, if a person is not injured or slightly injured in a traffic accident, and the vehicle is damaged When it is not big, after the relevant personnel (such as the car owner or passerby) collect and upload the traffic accident scene information, the car owner concerned can leave without having to stay on the scene, so as to avoid causing traffic congestion or shortening the congestion time. When a traffic accident occurs on a section where traffic monitoring equipment such as a camera is installed, the traffic monitoring equipment automatically collects and reports the traffic accident scene information, and the car owner involved may not even need to stay on the scene. In this way, road congestion caused by traffic accidents is avoided. This is of great significance to the traffic conditions of roads with heavy traffic and roads during peak periods.
- the generated traffic accident liability determination information may be further processed, such as storing, publishing, and sending to the parties involved in the traffic accident.
- the embodiment of the present application proposes a traffic accident processing method flow shown in FIG. 2.
- FIG. 2 is a second example flowchart of a traffic accident processing method provided by an embodiment of the present application. As shown in FIG. 2, in this embodiment, the method for processing a traffic accident may include the following steps:
- S202 Acquire a designated accident analysis model.
- the designated accident analysis model is obtained by training based on a deep learning algorithm and using known traffic accident responsibility information corresponding to known traffic accident scene information and known traffic accident scene information.
- S203 Based on the designated accident analysis model, generate traffic accident responsibility determination information according to traffic accident scene information, and execute steps S204, S205, and / or S206.
- the traffic department can store the traffic accident responsibility determination information in order to record the traffic accident and provide a basis for future inquiry.
- the designated media platform may be the official website of the transportation department.
- the traffic department can publish the information about the responsibility of the traffic accident to the official website for the relevant personnel to check.
- the designated media platform can further provide an interface for downloading traffic accident liability identification information, so that relevant personnel can easily download the traffic accident liability identification information from the designated media platform.
- the server can automatically send the traffic accident liability determination information to the mobile device of the traffic accident party. For example, a short message containing the contents of the traffic accident liability determination information is sent to the party's mobile phone. In this way, the parties to a traffic accident can be informed of the results of the traffic accident in a timely manner, and the parties to a traffic accident are not required to go to the relevant departments to collect them, saving time for both parties and the transportation department.
- the parties to a traffic accident can actively request to download the traffic accident liability determination information through the client without having to go to the relevant department to collect it, saving the time of the parties.
- step S204, step S205, and step S206 may be performed instead of all of them.
- S204, S205, and S206 are only step numbers, and do not indicate the execution order of the steps. The embodiment of the present application does not limit the execution order of steps S204, S205, and S206.
- the server When the server receives traffic accident scene information, it indicates that a traffic accident has occurred. At this time, the server can actively report to a designated agency (such as an insurance company), which further reduces user operations and saves time for the user.
- a designated agency such as an insurance company
- the embodiment of the present application proposes a traffic accident processing method flow shown in FIG. 3.
- FIG. 3 is a third exemplary flowchart of a traffic accident processing method according to an embodiment of the present application. As shown in FIG. 3, in this embodiment, the method for handling a traffic accident may include the following steps:
- the specified accident analysis model is based on a deep learning algorithm and is trained by using known traffic accident scene information and known traffic accident responsibility information corresponding to the known traffic accident scene information.
- the designated agency may be an insurance company, a transportation department, and so on.
- the server automatically reports to the designated agency after receiving the information of the traffic accident scene, and does not require the relevant personnel to call the designated agency to report the case, which facilitates the parties involved in the accident and saves processing time for the parties. In this way, the length of stay of the parties at the scene is further reduced, and the probability of traffic congestion is reduced.
- the traffic accident scene information includes the location of the traffic accident; analyzing the traffic accident scene information to determine the corresponding designated agency includes: determining that the traffic accident location has jurisdiction according to the traffic accident location The right traffic management agency; or, according to the location of the traffic accident, obtain first aid resources of all hospitals within the designated area centered on the location of the traffic accident; according to the first aid resources, select free from all hospitals within the designated area
- the hospital with emergency resources is the candidate hospital; from the candidate hospitals, the hospital closest to the location where the traffic accident occurred is determined as the target hospital, and the target hospital is the designated institution.
- the method may further include: sending an emergency resource retrieval instruction to the target hospital, the emergency resource retrieval instruction is used to retrieve the ambulance and medical resources of the target hospital, and the emergency resource retrieval instruction is Includes address information for the location of the traffic accident.
- the embodiment of the present application proposes a traffic accident processing method flow shown in FIG. 4.
- FIG. 4 is a fourth exemplary flowchart of a traffic accident processing method provided by an embodiment of the present application. As shown in FIG. 4, in this embodiment, the method for handling a traffic accident may include the following steps:
- S401 Receive traffic scene information.
- the designated accident analysis model is obtained by training based on a deep learning algorithm and using known traffic accident responsibility information corresponding to known traffic accident scene information and known traffic accident scene information.
- S405 If rescue is needed, send a rescue request message to the designated rescue agency.
- the server can intelligently determine whether rescue is needed based on the traffic accident scene information, and actively send rescue request information to the designated rescue agency when rescue is needed, so that the designated rescue agency can rush to the scene to treat the injured as soon as possible. This not only reduces the number of rescue operations required for the parties, but also provides a solution for situations where the parties cannot automatically ask for assistance, ensuring that injured people can be treated as soon as possible.
- the designated rescue agency may be a hospital.
- the rescue request information may include information on the location of the accident.
- the traffic accident scene information includes traffic accident scene images; judging whether rescue is needed based on the traffic accident scene information may include: extracting a bloodstain image from the traffic accident scene image to obtain an extraction result; and according to the extraction result To determine whether there is blood on the scene of a traffic accident; if there is blood on the scene of a traffic accident, determine that rescue is needed; or, if there is no blood on the scene of a traffic accident, determine that rescue is not required.
- judging whether rescue is needed based on traffic accident scene information may include: judging whether the traffic accident scene information shows that the vehicle owner's stay time exceeds the specified duration; if the traffic accident scene information shows that the vehicle owner's stay time exceeds the specified duration, It is determined that rescue is required; or, if the information on the scene of the traffic accident shows that the owner's stay time does not exceed the specified time, it is determined that rescue is not required.
- the traffic accident processing method receives traffic accident scene information, obtains a specified accident analysis model trained based on a deep learning algorithm, and generates a traffic accident based on the traffic accident scene information based on the specified accident analysis model.
- Responsibility determination information can use the specified accident analysis model to automatically and intelligently generate traffic accident responsibility determination information, without the need for traffic police and insurance personnel to conduct on-site investigations, and no need for traffic accident personnel to wait on the scene, on the one hand saving labor costs and reducing accidents
- the processing time improves the speed of accident handling.
- the waiting time of traffic accident personnel is greatly reduced, so that traffic accident personnel without major disturbance can quickly leave the accident site, avoiding traffic congestion or shortening the congestion time.
- the embodiment of the present application provides a traffic accident processing system, and the traffic accident processing system is configured to execute the traffic accident processing method of the foregoing embodiment.
- FIG. 5 is a structural block diagram of a traffic accident processing system according to an embodiment of the present application. As shown in FIG. 5, in this embodiment, the traffic accident processing system includes:
- the receiving module 510 is configured to receive traffic accident scene information.
- the obtaining module 520 is configured to obtain a specified accident analysis model.
- the specified accident analysis model is based on a deep learning algorithm and is trained by using known traffic accident scene information and known traffic accident responsibility determination information corresponding to the known traffic accident scene information.
- a generating module 530 is configured to generate traffic accident liability identification information based on traffic accident scene information based on a designated accident analysis model.
- the system may further include: a storage module, configured to store traffic accident responsibility determination information to a specified storage location.
- the system may further include: a request receiving module for receiving a download request for downloading the traffic accident responsibility determination information from the client; and an obtaining module for responding to the download request from a specified storage location Acquiring traffic accident liability determination information; a first sending module is used to send traffic accident liability determination information to a client.
- system may further include a publishing module for publishing traffic accident responsibility determination information to a designated media platform.
- system may further include: a second sending module, configured to send traffic accident responsibility determination information to the mobile device of the traffic accident party.
- the system may further include: an analysis module configured to analyze the traffic accident scene information in response to receiving the traffic accident scene information to determine a corresponding designated agency; a report information generation module, It is used to generate report information; a third sending module is used to send report information to a designated agency.
- the traffic accident scene information includes the location where the traffic accident occurred; when the analysis module is used to analyze the traffic accident scene information to determine the corresponding designated agency, it can be used to: Location, determine the traffic management agency that has jurisdiction over the location of the traffic accident; or, based on the location of the traffic accident, obtain first aid resources for all hospitals within the designated area centered on the location of the traffic accident; Among all the hospitals in the area, a hospital having free emergency resources is selected as an alternative hospital. From the alternative hospitals, a hospital closest to the place where a traffic accident occurs is determined as a target hospital, and the target hospital is a designated institution.
- the acquisition module 520 when the acquisition module 520 is used to acquire a specified accident analysis model, the acquisition module 520 may be used to: acquire multiple sets of training data, each set of training data including known traffic accident scene information and the known traffic accident Known traffic accident responsibility identification information corresponding to the scene information; extract known traffic accident scene information from multiple sets of training data as input data for a designated accident analysis model; extract known traffic accident liability identification information from multiple sets of training data, As the output data of the specified accident analysis model; use the input data and output data to train the deep learning algorithm model to generate the specified accident analysis model.
- the system may further include: a determination module for determining whether rescue is needed based on traffic accident scene information; and a fourth sending module for sending a rescue request to a designated rescue agency if rescue is needed information.
- the traffic accident scene information includes a traffic accident scene image; when the judgment module is used to determine whether rescue is required based on the traffic accident scene information, it may be specifically used to: extract a bloodstain image from the traffic accident scene image To obtain the extraction result; to determine whether there is blood on the scene of the traffic accident according to the extraction result; if there is blood on the scene of the traffic accident, it is necessary to rescue;
- the judgment module when used to determine whether rescue is needed based on traffic accident scene information, it can be specifically used to: determine whether the traffic accident scene information shows that the vehicle owner's stay time exceeds the specified duration; if the traffic accident scene information If the vehicle owner stays longer than the specified time, it is determined that rescue is needed; or if the traffic accident scene information shows that the vehicle owner stays less than the specified time, it is determined that no rescue is needed.
- the traffic accident processing system receives traffic accident scene information, obtains a specified accident analysis model trained based on a deep learning algorithm, and generates traffic accident responsibility determination information based on the specified information of the accident accident scene based on the specified accident analysis model.
- the waiting time of traffic accident personnel is greatly reduced, so that the traffic accident personnel without major disturbance can quickly leave the scene of the accident, avoiding causing traffic congestion or shortening the congestion time.
- An embodiment of the present application provides a server.
- the server includes: a processor; a memory for storing processor-executable instructions; the processor is configured to: receive traffic accident scene information; obtain a designated accident analysis model; Based on a deep learning algorithm, trained using known traffic accident scene information and known traffic accident liability identification information corresponding to the known traffic accident scene information; based on a designated accident analysis model, generating traffic accident liability based on the traffic accident scene information Authorization information.
- FIG. 6 is a simplified block diagram of a server provided by an embodiment of the present application.
- the server 600 may include a processor 601 connected to one or more data storage tools, and the data storage tool may include a storage medium 606 and a memory unit 604.
- the server 600 may further include an input interface 605 and an output interface 607 for communicating with another device or system.
- the program code executed by the CPU of the processor 601 may be stored in the memory unit 604 or the storage medium 606.
- the processor 601 in the server 600 calls the program code stored in the memory unit 604 or the storage medium 606 and performs the following steps:
- the specified accident analysis model is based on a deep learning algorithm and is trained using known traffic accident responsibility information corresponding to known traffic accident scene information and known traffic accident scene information.
- the traffic accident responsibility identification information is generated based on the traffic accident scene information.
- the processor 601 may be further configured to: store the traffic accident responsibility determination information to a designated storage location.
- the processor 601 may be further configured to:
- the processor 601 may be further configured to:
- the processor 601 may be further configured to:
- the processor 601 may be further configured to:
- the traffic accident scene information includes the location where the traffic accident occurred; the processor 601 may also be configured to:
- the location of the traffic accident obtain the first-aid resources of all hospitals within the designated area centered on the location of the traffic accident; according to the first-aid resources, select the hospital with free emergency resources from all the hospitals within the designated area as the first-aid resources.
- Candidate hospitals From among the alternative hospitals, determine the hospital closest to the location where the traffic accident occurred as the target hospital and the target hospital as the designated institution.
- the processor 601 may be further configured to:
- each set of training data including known traffic accident scene information and known traffic accident responsibility identification information corresponding to the known traffic accident scene information;
- the processor 601 may be further configured to:
- rescue If rescue is needed, send a rescue request message to the designated rescue agency.
- the traffic accident scene information includes a traffic accident scene image; the processor 601 may also be configured to:
- the processor 601 may be further configured to:
- the traffic accident scene information shows that the vehicle owner stays longer than the specified time, it is determined that rescue is needed; or if the traffic accident scene information shows that the vehicle owner stays time does not exceed the specified time, it is determined that rescue is not required.
- An embodiment of the present application provides a computer non-volatile readable storage medium.
- the computer non-volatile readable storage medium stores a computer program.
- the computer program is executed by a processor, the following steps are implemented:
- the specified accident analysis model is based on a deep learning algorithm and is trained by using known traffic accident responsibility information corresponding to known traffic accident scene information and known traffic accident scene information.
- the traffic accident responsibility identification information is generated based on the traffic accident scene information.
- the following steps are also implemented: storing the traffic accident liability determination information to a designated storage location.
- the traffic accident scene information includes the location where the traffic accident occurred; when the computer program is executed by the processor, the following steps are also implemented:
- the location of the traffic accident obtain the first-aid resources of all hospitals within the designated area centered on the location of the traffic accident; according to the first-aid resources, select the hospital with free emergency resources from all hospitals within the designated area
- Candidate hospitals From among the alternative hospitals, determine the hospital closest to the location where the traffic accident occurred as the target hospital and the target hospital as the designated institution.
- each set of training data including known traffic accident scene information and known traffic accident responsibility identification information corresponding to the known traffic accident scene information;
- rescue If rescue is needed, send a rescue request message to the designated rescue agency.
- traffic accident scene information includes traffic accident scene images; when a computer program is executed by a processor, the following steps are also implemented:
- the traffic accident scene information shows that the vehicle owner stays longer than the specified time, it is determined that rescue is needed; or if the traffic accident scene information shows that the vehicle owner stays time does not exceed the specified time, it is determined that rescue is not required.
Landscapes
- Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Development Economics (AREA)
- Educational Administration (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Traffic Control Systems (AREA)
Abstract
本申请实施例提供了一种交通事故处理方法、系统及服务器。本申请实施例通过接收交通事故现场信息，获取指定事故分析模型，基于指定事故分析模型，根据交通事故现场信息，生成交通事故责任认定信息，能够利用指定事故分析模型，自动智能地生成交通事故责任认定信息，不需要交警和保险人员到现场勘查，不需要交通事故人员在现场等待，一方面节约了人力成本，减少了事故处理时间，提高了事故处理速度，另一方面大大减少了交通事故人员的等待时间，使得没有大碍的交通事故人员能够快速离开事故现场，避免造成交通拥堵或者缩短拥堵时间，在一定程度上解决了现有技术中交通事故的处理流程存在人力成本高且耗时较多的问题。
Description
本申请要求于2018年6月28日提交中国专利局、申请号为201810690318.5、发明名称为“交通事故处理方法、系统及服务器”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
本申请涉及通信技术领域，尤其涉及一种交通事故处理方法、系统及服务器。
当前，随着我国国内汽车数量的持续增长，公路的交通压力越来越大。例如，大中型城市的市内道路经常发生拥堵，特别是上下班、节假日等高峰期。当发生交通事故时，拥堵状况就会更加严重。
目前，发生交通事故时，处理流程是：交通事故的相关人员进行报案，交通部门和保险公司接到报案后，指派人员到交通事故现场进行现场勘查(比如拍照等)，然后由交警当场判定事故责任，开具交通事故责任认定书。或者，交警后续通过调取事故现场的监控录像，并结合现场勘查信息，在交通事故发生数日后判定事故责任，开具交通事故责任认定书。
在上述处理过程中，事故车辆的停留本身就造成了事故现场路段的拥堵，而交警车辆、保险公司车辆的到来会进一步加剧事故现场路段的拥堵。此过程中，交通部门和保险公司都需要指派人员亲临现场，人力成本较大。并且，由于事故是由人工现场勘查，因此事故处理时间较长，相关人员需要耗费较多的等待时间。
申请内容
本申请实施例提供了一种交通事故处理方法、系统及服务器，用以解决现有技术中交通事故的处理流程存在人力成本高且耗时较多的问题。
第一方面，本申请实施例提供一种交通事故处理方法，所述方法包括：
接收交通事故现场信息；
获取指定事故分析模型，所述指定事故分析模型是基于深度学习算法，利用已知交通事故现场信息和所述已知交通事故现场信息 对应的已知交通事故责任认定信息训练得到的；
基于所述指定事故分析模型，根据所述交通事故现场信息，生成交通事故责任认定信息。
第二方面，本申请实施例提供一种交通事故处理系统，所述系统包括：
接收模块，用于接收交通事故现场信息；
获取模块，用于获取指定事故分析模型，所述指定事故分析模型是基于深度学习算法，利用已知交通事故现场信息和所述已知交通事故现场信息对应的已知交通事故责任认定信息训练得到的；
生成模块，用于基于指所述定事故分析模型，根据所述交通事故现场信息，生成交通事故责任认定信息。
第三方面，本申请实施例提供一种服务器，所述服务器包括：
处理器；
用于存储所述处理器可执行指令的存储器；
所述处理器被配置为：
接收交通事故现场信息；
获取指定事故分析模型，所述指定事故分析模型是基于深度学习算法，利用已知交通事故现场信息和所述已知交通事故现场信息对应的已知交通事故责任认定信息训练得到的；
基于所述指定事故分析模型，根据所述交通事故现场信息，生成交通事故责任认定信息。
第四方面，本申请实施例提供一种计算机非易失性可读存储介质，所述计算机非易失性可读存储介质存储有计算机程序，所述计算机程序被处理器执行时实现如下步骤：
接收交通事故现场信息；
获取指定事故分析模型，所述指定事故分析模型是基于深度学习算法，利用已知交通事故现场信息和所述已知交通事故现场信息对应的已知交通事故责任认定信息训练得到的；
基于所述指定事故分析模型，根据所述交通事故现场信息，生成交通事故责任认定信息。
本申请实施例具有以下有益效果：
本申请实施例，通过接收交通事故现场信息，获取基于深度学习算法训练得到的指定事故分析模型，基于指定事故分析模型，根据交通事故现场信息，生成交通事故责任认定信息，能够利用指定事故分析模型，自动智能地生成交通事故责任认定信息，不需要交 警和保险人员到现场勘查，不需要交通事故人员在现场等待，一方面节约了人力成本，减少了事故处理时间，提高了事故处理速度，另一方面大大减少了交通事故人员的等待时间，使得没有大碍的交通事故人员能够快速离开事故现场，避免造成交通拥堵或者缩短拥堵时间。
图1为本申请实施例提供的交通事故处理方法的第一流程示例图。
图2为本申请实施例提供的交通事故处理方法的第二流程示例图。
图3为本申请实施例提供的交通事故处理方法的第三流程示例图。
图4为本申请实施例提供的交通事故处理方法的第四流程示例图。
图5为本申请实施例提供的交通事故处理系统的结构框图。
图6是本申请实施例提供的服务器的简化框图。
为了更好的理解本申请的技术方案，下面结合附图对本申请实施例进行详细描述。
实施例一
本申请实施例提供一种交通事故处理方法。本申请实施例提供的交通事故处理方法可以执行在交通部门的服务器上。
图1为本申请实施例提供的交通事故处理方法的第一流程示例图。如图1所示，本实施例中，交通事故处理方法可以包括以下步骤：
S101，接收交通事故现场信息。
S102，获取指定事故分析模型，指定事故分析模型是基于深度学习算法，利用已知交通事故现场信息和已知交通事故现场信息对应的已知交通事故责任认定信息训练得到的。
S103，基于指定事故分析模型，根据交通事故现场信息，生成交通事故责任认定信息。
其中，交通事故现场信息可以包括交通事故现场的图片、视频 等。例如，发生交通事故时，事故现场环境、涉事车辆的现场照片、监控录像等都可以作为交通事故现场信息。
其中，步骤S101中接收的交通事故现场信息的来源可以是交通事故现场所在路段的交通监控设备，也可以是涉事车主的手机、平板等个人手持设备。
例如，当发生交通事故时，交通事故现场所在路段的交通监控设备可以实时采集现场图片，并将该现场图片连同存储的交通事故发生时间段的视频一同上传至交通部门的服务器。服务器即可接收到交通事故现场信息。交通监控设备采集、上传图片和/或视频的操作可以是基于自身的判断程序判断结果主动执行的，也可以是在接收到交通部门后台系统发出的指令而被动触发的。
相应地，接收交通事故现场信息的方式可以是：接收交通事故现场所在路段的交通监控设备采集的交通事故现场信息。
例如，当发生交通事故时，涉事车主(也可以称为交通事故当事人)、路人等可以通过手机、平板等个人手持设备拍摄现场图片或现场视频，然后通过网络将现场图片和/或现场视频发送给交通部门的服务器。
所述的服务器可以设置用于接收涉事车主、路人的个人手持设备上传信息的专用接口。例如，该接口可以是交通部门的官方网站的数据上传接口，还可以是个人手持设备上安装的应用APP(Application)，或者该应用上的接口子程序。
相应地，接收交通事故现场信息的方式可以是：接收个人手持设备上传的交通事故现场信息。
需要说明的是，上述仅是接收交通事故现场信息的方式的举例，并不用于对接收交通事故现场信息的方式进行限定，在应用中，可以根据实际的情况和需求来设置接收交通事故现场信息的具体方式。
可见，基于步骤S101，交通部门的后台服务器能够方便快捷地获取到交通事故现场信息，而不必指派人员到交通事故现场进行人工现场勘查，既有助于缩短事故的处理时间，提高事故处理效率，又节约了人力成本。
其中，步骤S102中的指定事故分析模型可以是预先训练好的。在一个示例性的实现过程中，获取指定事故分析模型，可以包括：获取多组训练数据，每组训练数据包括已知交通事故现场信息和与该已知交通事故现场信息对应的已知交通事故责任认定信息；提取多组训练数据中的已知交通事故现场信息，作为指定事故分析模型 的输入数据；提取多组训练数据中的已知交通事故责任认定信息，作为指定事故分析模型的输出数据；利用输入数据和输出数据对深度学习算法模型进行训练，生成指定事故分析模型。
这样，通过大量已知的交通事故现场信息，以及与这些已知交通事故现场信息对应的已知交通事故责任认定信息，基于深度学习算法，就可以训练得到指定事故分析模型。
需要说明的是，训练数据的组数越多，训练得到的指定事故分析模型就越精确。
下面举例说明指定事故分析模型的获取过程。假设深度学习算法模型采用卷积神经网络模型，指定事故分析模型的获取过程可以是：将已经发生的一万组已知交通事故现场信息和与该已知交通事故现场信息对应的已知交通事故责任认定信息作为训练数据，为卷积神经网络模型设定若干参数，这些参数在训练之前为未知数，将训练数据中的已知交通事故现场信息作为卷积神经网络模型的输入，将训练数据中的已知交通事故责任认定信息作为卷积神经网络模型的输出，求解出所述若干参数的参数值，此时，这些参数的参数值为已知数，这些参数值确定了一种具体的卷积神经网络模型，该卷积神经网络模型即可作为指定事故分析模型。
步骤S102中，基于指定事故分析模型，根据交通事故现场信息，生成交通事故责任认定信息，可以包括：将交通事故现场信息作为输入数据输入到指定事故分析模型中；获取指定事故分析模型的输出数据，将该输出数据作为交通事故责任认定信息。
其中，交通事故责任认定信息可以是交通事故责任认定书。
其中，交通事故责任认定信息的内容可以包括：
道路交通事故当事人、车辆、道路和交通环境等基本情况；
道路交通事故发生经过；
道路交通事故证据及事故形成原因的分析；
当事人导致道路交通事故的过错及责任或者意外原因；
作出道路交通事故认定的公安机关交通管理部门名称和日期。
需要说明的是，在具体应用中，可以根据需要或实际情况对交通事故责任认定信息所包含的具体内容进行调整，比如增加、减少或修改相应的内容等。
通过步骤S102，交通部门的后台服务器可以自动智能地对交通事故进行分析判断，进而确定交通事故的责任。这样，就不再需要交警赶到交通事故现场，再人工判定交通事故的责任，从而减少了 事故处理时间，提高了事故处理速度。
根据步骤S101和步骤S102可知，本申请实施例根据远程采集的交通事故现场信息就可以自动智能地为交通事故定责，因此，交通事故中，如果人员没有受伤或受轻伤，并且车辆的损坏不大时，在相关人员(例如涉事车主或路人)采集并上传交通事故现场信息后，涉事车主就可以自行离开，而不必留在现场，这样就可以避免造成交通拥堵或者缩短拥堵时间。当交通事故发生在安装有摄像头等交通监控设备的路段时，交通监控设备自动采集并上报交通事故现场信息，涉事车主甚至可以不用在现场停留。这样，就避免了因交通事故而造成道路拥堵。这对于车流量较大的道路以及高峰期的道路的交通状况，具有重大意义。
在本申请其他实施例中，在生成交通事故责任认定信息之后，可以进一步对生成的交通事故责任认定信息进行处理，比如存储、发布、发给交通事故的当事人等等。
基于此，本申请实施例提出了图2所示的交通事故处理方法流程。
图2为本申请实施例提供的交通事故处理方法的第二流程示例图。如图2所示，本实施例中，交通事故处理方法可以包括以下步骤：
S201，接收交通事故现场信息。
S202，获取指定事故分析模型，指定事故分析模型是基于深度学习算法，利用已知交通事故现场信息和已知交通事故现场信息对应的已知交通事故责任认定信息训练得到的。
S203，基于指定事故分析模型，根据交通事故现场信息，生成交通事故责任认定信息，执行步骤S204、S205和/或S206。
S204，将交通事故责任认定信息存储到指定存储位置，执行步骤S207。
S205，将交通事故责任认定信息发布到指定媒介平台，结束。
S206，向交通事故当事人的移动设备发送交通事故责任认定信息，结束。
S207，接收来自客户端的下载交通事故责任认定信息的下载请求。
S208，响应于下载请求，从指定存储位置获取交通事故责任认定信息。
S209，向客户端发送交通事故责任认定信息。
通过步骤S204，交通部门可以对交通事故责任认定信息进行存储，以便对交通事故进行备案，以及为以后的查询提供基础。
步骤S205中，指定媒介平台可以是交通部门的官方网站。通过步骤S205，交通部门可以将交通事故责任认定信息发布到官方网站上，以供相关人员查阅。
其中，指定媒介平台还可以进一步提供下载交通事故责任认定信息的接口，以便相关人员能够从指定媒介平台方便地下载交通事故责任认定信息。
通过步骤S206，服务器在生成交通事故责任认定信息后，可以自动向交通事故当事人的移动设备发送交通事故责任认定信息。例如向当事人的手机发送含有交通事故责任认定信息的内容的短信息。这样，可以让交通事故的当事人及时获知交通事故的处理结果，不需要交通事故的当事人到相关部门去领取，节省了当事人和交通部门双方的时间。
通过步骤S207～S209，交通事故的当事人可以通过客户端主动请求下载交通事故责任认定信息，而不必到相关部门去领取，节省了当事人的时间。
需要说明的是，在其他实施例中，也可以执行步骤S204、步骤S205、步骤S206中的任意一个或两个步骤，而不是全部执行。并且，S204、S205、S206仅为步骤编号，并不表示步骤的执行顺序。本申请实施例不对步骤S204、步骤S205、步骤S206的执行顺序进行限定。
当服务器接收到交通事故现场信息时，说明有交通事故发生，此时服务器可以主动向指定机构(比如保险公司)报案，这样就进一步减少了用户的操作，为用户节省了时间。
基于此，本申请实施例提出了图3所示的交通事故处理方法流程。
图3为本申请实施例提供的交通事故处理方法的第三流程示例图。如图3所示，本实施例中，交通事故处理方法可以包括以下步骤：
S301，接收交通事故现场信息。
S302，获取指定事故分析模型，指定事故分析模型是基于深度学习算法，利用已知交通事故现场信息和已知交通事故现场信息对应的已知交通事故责任认定信息训练得到的。
S303，基于指定事故分析模型，根据交通事故现场信息，生成 交通事故责任认定信息，
S304，响应于接收到交通事故现场信息，对交通事故现场信息进行分析，以确定对应的指定机构。
S305，生成报案信息。
S306，将报案信息发送给指定机构。
其中，指定机构可以是保险公司、交通部门等。
通过步骤S304至步骤S306，服务器在接收到交通事故现场信息后，自动向指定机构报案，不需要相关人员再打电话给指定机构报案，方便了事故当事人，为当事人节省了处理时间。这样，就进一步减少了当事人在现场的停留时间，降低了发生交通拥堵的概率。
在一个示例性的实现过程中，交通事故现场信息包括交通事故发生地段；对交通事故现场信息进行分析，以确定对应的指定机构，包括：根据交通事故发生地段，确定对交通事故发生地段具有管辖权的交通管理机构；或者，根据交通事故发生地段，获取以交通事故发生地段为中心的指定区域范围内的所有医院的急救资源；根据急救资源，从指定区域范围内的所有医院中选择有空闲急救资源的医院，作为备选医院；从备选医院中，确定距离交通事故发生地段最近的医院，作为目标医院，以目标医院作为指定机构。
在一个示例性的实现过程中，所述方法还可以包括：向目标医院发送急救资源调取指令，急救资源调取指令用于调取目标医院的救护车和医疗资源，急救资源调取指令中包括交通事故发生地段的地址信息。
在一些交通事故中，可能会出现人员受伤较严重、无法自行离开现场而需要救援的情况，此时，不仅需要向交通部门和保险公司报案，还需要请求救援。本申请实施例中，可以对是否需要救援的情况进行判断，然后根据判断结果确定是否需要救援。
基于此，本申请实施例提出了图4所示的交通事故处理方法流程。
图4为本申请实施例提供的交通事故处理方法的第四流程示例图。如图4所示，本实施例中，交通事故处理方法可以包括以下步骤：
S401，接收交通事故现场信息。
S402，获取指定事故分析模型，指定事故分析模型是基于深度学习算法，利用已知交通事故现场信息和已知交通事故现场信息对应的已知交通事故责任认定信息训练得到的。
S403，基于指定事故分析模型，根据交通事故现场信息，生成交通事故责任认定信息，
S404，根据交通事故现场信息，判断是否需要救援。
S405，若需要救援，向指定救援机构发送救援请求信息。
通过步骤S404和S405，服务器可以根据交通事故现场信息智能地判断是否需要救援，并在需要救援时主动向指定救援机构发送救援请求信息，以便指定救援机构能够尽快赶赴现场对受伤人员进行救治。这不仅为当事人减少了请求救援的操作，还为当事人无法自动求援的情况提供了解决方案，保障受伤人员能够尽快得到救治。
其中，指定救援机构可以是医院。
其中，救援请求信息中可以包含事故地点信息。
在一个示例性的实现过程中，交通事故现场信息包括交通事故现场图像；根据交通事故现场信息，判断是否需要救援，可以包括：从交通事故现场图像中提取血迹图像，得到提取结果；根据提取结果，判断交通事故现场是否有血迹；若交通事故现场有血迹，判定需要救援；或者，若交通事故现场没有血迹，判定不需要救援。
在一个示例性的实现过程中，根据交通事故现场信息，判断是否需要救援，可以包括：判断交通事故现场信息是否显示车主停留时间超过指定时长；若交通事故现场信息显示车主停留时间超过指定时长，判定需要救援；或者，若交通事故现场信息显示车主停留时间未超过指定时长，判定不需要救援。
在本申请其他实施例中，也可以结合上述两种方式综合判断是否需要救援，或者根据除血迹、停留时间以外的其他因素判断是否需要救援，本申请对此不进行限定。
综上可见，本申请实施例提供的交通事故处理方法，通过接收交通事故现场信息，获取基于深度学习算法训练得到的指定事故分析模型，基于指定事故分析模型，根据交通事故现场信息，生成交通事故责任认定信息，能够利用指定事故分析模型，自动智能地生成交通事故责任认定信息，不需要交警和保险人员到现场勘查，不需要交通事故人员在现场等待，一方面节约了人力成本，减少了事故处理时间，提高了事故处理速度，另一方面大大减少了交通事故人员的等待时间，使得没有大碍的交通事故人员能够快速离开事故现场，避免造成交通拥堵或者缩短拥堵时间。
实施例二
本申请实施例提供一种交通事故处理系统，该交通事故处理系 统用于执行前述实施例一种的交通事故处理方法。
图5为本申请实施例提供的交通事故处理系统的结构框图。如图5所示，本实施例中，交通事故处理系统包括：
接收模块510，用于接收交通事故现场信息。
获取模块520，用于获取指定事故分析模型，指定事故分析模型是基于深度学习算法，利用已知交通事故现场信息和已知交通事故现场信息对应的已知交通事故责任认定信息训练得到的。
生成模块530，用于基于指定事故分析模型，根据交通事故现场信息，生成交通事故责任认定信息。
在一个示例性的实现过程中，所述系统还可以包括：存储模块，用于将交通事故责任认定信息存储到指定存储位置。
在一个示例性的实现过程中，所述系统还可以包括：请求接收模块，用于接收来自客户端的下载交通事故责任认定信息的下载请求；获取模块，用于响应于下载请求，从指定存储位置获取交通事故责任认定信息；第一发送模块，用于向客户端发送交通事故责任认定信息。
在一个示例性的实现过程中，所述系统还可以包括：发布模块，用于将交通事故责任认定信息发布到指定媒介平台。
在一个示例性的实现过程中，所述系统还可以包括：第二发送模块，用于向交通事故当事人的移动设备发送交通事故责任认定信息。
在一个示例性的实现过程中，所述系统还可以包括：分析模块，用于响应于接收到交通事故现场信息，对交通事故现场信息进行分析，以确定对应的指定机构；报案信息生成模块，用于生成报案信息；第三发送模块，用于将报案信息发送给指定机构。
在一个示例性的实现过程中，所述交通事故现场信息包括交通事故发生地段；分析模块在用于对交通事故现场信息进行分析，以确定对应的指定机构时，可以用于：根据交通事故发生地段，确定对交通事故发生地段具有管辖权的交通管理机构；或者，根据交通事故发生地段，获取以交通事故发生地段为中心的指定区域范围内的所有医院的急救资源；根据急救资源，从指定区域范围内的所有医院中选择有空闲急救资源的医院，作为备选医院；从所述备选医院中，确定距离交通事故发生地段最近的医院，作为目标医院，以目标医院作为指定机构。
在一个示例性的实现过程中，获取模块520在用于获取指定事 故分析模型时，可以用于：获取多组训练数据，每组训练数据包括已知交通事故现场信息和与该已知交通事故现场信息对应的已知交通事故责任认定信息；提取多组训练数据中的已知交通事故现场信息，作为指定事故分析模型的输入数据；提取多组训练数据中的已知交通事故责任认定信息，作为指定事故分析模型的输出数据；利用输入数据和输出数据对深度学习算法模型进行训练，生成指定事故分析模型。
在一个示例性的实现过程中，所述系统还可以包括：判断模块，用于根据交通事故现场信息，判断是否需要救援；第四发送模块，用于若需要救援，向指定救援机构发送救援请求信息。
在一个示例性的实现过程中，交通事故现场信息包括交通事故现场图像；判断模块在用于根据交通事故现场信息，判断是否需要救援时，可以具体用于：从交通事故现场图像中提取血迹图像，得到提取结果；根据提取结果，判断交通事故现场是否有血迹；若交通事故现场有血迹，判定需要救援；或者，若交通事故现场没有血迹，判定不需要救援。
在一个示例性的实现过程中，判断模块在用于根据交通事故现场信息，判断是否需要救援时，可以具体用于：判断交通事故现场信息是否显示车主停留时间超过指定时长；若交通事故现场信息显示车主停留时间超过指定时长，判定需要救援；或者，若交通事故现场信息显示车主停留时间未超过指定时长，判定不需要救援。
本申请实施例提供的交通事故处理系统，通过接收交通事故现场信息，获取基于深度学习算法训练得到的指定事故分析模型，基于指定事故分析模型，根据交通事故现场信息，生成交通事故责任认定信息，能够利用指定事故分析模型，自动智能地生成交通事故责任认定信息，不需要交警和保险人员到现场勘查，不需要交通事故人员在现场等待，一方面节约了人力成本，减少了事故处理时间，提高了事故处理速度，另一方面大大减少了交通事故人员的等待时间，使得没有大碍的交通事故人员能够快速离开事故现场，避免造成交通拥堵或者缩短拥堵时间。
实施例三
本申请实施例提供一种服务器，该服务器包括：处理器；用于存储处理器可执行指令的存储器；处理器被配置为：接收交通事故现场信息；获取指定事故分析模型，指定事故分析模型是基于深度学习算法，利用已知交通事故现场信息和所述已知交通事故现场信 息对应的已知交通事故责任认定信息训练得到的；基于指定事故分析模型，根据交通事故现场信息，生成交通事故责任认定信息。
图6是本申请实施例提供的服务器的简化框图。请参见图6，该服务器600可以包括与一个或多个数据存储工具连接的处理器601，该数据存储工具可以包括存储介质606和内存单元604。服务器600还可以包括输入接口605和输出接口607，用于与另一装置或系统进行通信。被处理器601的CPU执行的程序代码可存储在内存单元604或存储介质606中。
服务器600中的处理器601调用存储在内存单元604或存储介质606的程序代码，执行下面各步骤：
接收交通事故现场信息；
获取指定事故分析模型，指定事故分析模型是基于深度学习算法，利用已知交通事故现场信息和已知交通事故现场信息对应的已知交通事故责任认定信息训练得到的；
基于指定事故分析模型，根据交通事故现场信息，生成交通事故责任认定信息。
在一个示例性的实现过程中，处理器601还可以被配置为：将交通事故责任认定信息存储到指定存储位置。
在一个示例性的实现过程中，处理器601还可以被配置为：
接收来自客户端的下载交通事故责任认定信息的下载请求；
响应于下载请求，从指定存储位置获取交通事故责任认定信息；
向客户端发送交通事故责任认定信息。
在一个示例性的实现过程中，处理器601还可以被配置为：
将交通事故责任认定信息发布到指定媒介平台。
在一个示例性的实现过程中，处理器601还可以被配置为：
向交通事故当事人的移动设备发送交通事故责任认定信息。
在一个示例性的实现过程中，处理器601还可以被配置为：
响应于接收到交通事故现场信息，对交通事故现场信息进行分析，以确定对应的指定机构；
生成报案信息；
将报案信息发送给指定机构。
在一个示例性的实现过程中，交通事故现场信息包括交通事故发生地段；处理器601还可以被配置为：
根据交通事故发生地段，确定对交通事故发生地段具有管辖权的交通管理机构；或者，
根据交通事故发生地段，获取以交通事故发生地段为中心的指定区域范围内的所有医院的急救资源；根据急救资源，从所述指定区域范围内的所有医院中选择有空闲急救资源的医院，作为备选医院；从备选医院中，确定距离交通事故发生地段最近的医院，作为目标医院，以目标医院作为指定机构。
在一个示例性的实现过程中，处理器601还可以被配置为：
获取多组训练数据，每组训练数据包括已知交通事故现场信息和与该已知交通事故现场信息对应的已知交通事故责任认定信息；
提取多组训练数据中的已知交通事故现场信息，作为指定事故分析模型的输入数据；
提取多组训练数据中的已知交通事故责任认定信息，作为指定事故分析模型的输出数据；
利用输入数据和输出数据对深度学习算法模型进行训练，生成指定事故分析模型。
在一个示例性的实现过程中，处理器601还可以被配置为：
根据交通事故现场信息，判断是否需要救援；
若需要救援，向指定救援机构发送救援请求信息。
在一个示例性的实现过程中，交通事故现场信息包括交通事故现场图像；处理器601还可以被配置为：
从交通事故现场图像中提取血迹图像，得到提取结果；
根据提取结果，判断交通事故现场是否有血迹；
若交通事故现场有血迹，判定需要救援；或者，若交通事故现场没有血迹，判定不需要救援。
在一个示例性的实现过程中，处理器601还可以被配置为：
判断交通事故现场信息是否显示车主停留时间超过指定时长；
若交通事故现场信息显示车主停留时间超过指定时长，判定需要救援；或者，若交通事故现场信息显示车主停留时间未超过指定时长，判定不需要救援。
实施例四
本申请实施例提供一种计算机非易失性可读存储介质，该计算机非易失性可读存储介质存储有计算机程序，该计算机程序被处理器执行时实现如下步骤：
接收交通事故现场信息；
获取指定事故分析模型，指定事故分析模型是基于深度学习算法，利用已知交通事故现场信息和已知交通事故现场信息对应的已 知交通事故责任认定信息训练得到的；
基于指定事故分析模型，根据交通事故现场信息，生成交通事故责任认定信息。
在一个示例性的实现过程中，计算机程序被处理器执行时还实现如下步骤：将交通事故责任认定信息存储到指定存储位置。
在一个示例性的实现过程中，计算机程序被处理器执行时还实现如下步骤：
接收来自客户端的下载交通事故责任认定信息的下载请求；
响应于下载请求，从指定存储位置获取交通事故责任认定信息；
向客户端发送交通事故责任认定信息。
在一个示例性的实现过程中，计算机程序被处理器执行时还实现如下步骤：
将交通事故责任认定信息发布到指定媒介平台。
在一个示例性的实现过程中，计算机程序被处理器执行时还实现如下步骤：
向交通事故当事人的移动设备发送交通事故责任认定信息。
在一个示例性的实现过程中，计算机程序被处理器执行时还实现如下步骤：
响应于接收到交通事故现场信息，对交通事故现场信息进行分析，以确定对应的指定机构；
生成报案信息；
将报案信息发送给指定机构。
在一个示例性的实现过程中，交通事故现场信息包括交通事故发生地段；计算机程序被处理器执行时还实现如下步骤：
根据交通事故发生地段，确定对交通事故发生地段具有管辖权的交通管理机构；或者，
根据交通事故发生地段，获取以交通事故发生地段为中心的指定区域范围内的所有医院的急救资源；根据急救资源，从所述指定区域范围内的所有医院中选择有空闲急救资源的医院，作为备选医院；从备选医院中，确定距离交通事故发生地段最近的医院，作为目标医院，以目标医院作为指定机构。
在一个示例性的实现过程中，计算机程序被处理器执行时还实现如下步骤：
获取多组训练数据，每组训练数据包括已知交通事故现场信息和与该已知交通事故现场信息对应的已知交通事故责任认定信息；
提取多组训练数据中的已知交通事故现场信息，作为指定事故分析模型的输入数据；
提取多组训练数据中的已知交通事故责任认定信息，作为指定事故分析模型的输出数据；
利用输入数据和输出数据对深度学习算法模型进行训练，生成指定事故分析模型。
在一个示例性的实现过程中，计算机程序被处理器执行时还实现如下步骤：
根据交通事故现场信息，判断是否需要救援；
若需要救援，向指定救援机构发送救援请求信息。
在一个示例性的实现过程中，交通事故现场信息包括交通事故现场图像；计算机程序被处理器执行时还实现如下步骤：
从交通事故现场图像中提取血迹图像，得到提取结果；
根据提取结果，判断交通事故现场是否有血迹；
若交通事故现场有血迹，判定需要救援；或者，若交通事故现场没有血迹，判定不需要救援。
在一个示例性的实现过程中，计算机程序被处理器执行时还实现如下步骤：
判断交通事故现场信息是否显示车主停留时间超过指定时长；
若交通事故现场信息显示车主停留时间超过指定时长，判定需要救援；或者，若交通事故现场信息显示车主停留时间未超过指定时长，判定不需要救援。
以上所述仅为本申请的较佳实施例而已，并不用以限制本申请，凡在本申请的精神和原则之内，所做的任何修改、等同替换、改进等，均应包含在本申请保护的范围之内。
Claims (20)
- 一种交通事故处理方法，其特征在于，所述方法包括：接收交通事故现场信息；获取指定事故分析模型，所述指定事故分析模型是基于深度学习算法，利用已知交通事故现场信息和所述已知交通事故现场信息对应的已知交通事故责任认定信息训练得到的；基于所述指定事故分析模型，根据所述交通事故现场信息，生成交通事故责任认定信息。
- 根据权利要求1所述的方法，其特征在于，所述方法还包括：接收来自客户端的下载所述交通事故责任认定信息的下载请求；响应于所述下载请求，从所述指定存储位置获取所述交通事故责任认定信息；向所述客户端发送所述交通事故责任认定信息。
- 根据权利要求1所述的方法，其特征在于，所述方法还包括：响应于接收到所述交通事故现场信息，对所述交通事故现场信息进行分析，以确定对应的指定机构；生成报案信息；将所述报案信息发送给所述指定机构。
- 根据权利要求1所述的方法，其特征在于，所述方法还包括：根据所述交通事故现场信息，判断是否需要救援；若需要救援，向指定救援机构发送救援请求信息。
- 根据权利要求4所述的方法，其特征在于，根据所述交通事故现场信息，判断是否需要救援，包括：所述交通事故现场信息包括交通事故现场图像；从所述交通事故现场图像中提取血迹图像，得到提取结果；根据所述提取结果，判断交通事故现场是否有血迹；若所述交通事故现场有血迹，判定需要救援；或者，若所述交通事故现场没有血迹，判定不需要救援。
- 一种交通事故处理系统，其特征在于，所述系统包括：接收模块，用于接收交通事故现场信息；获取模块，用于获取指定事故分析模型，所述指定事故分析模型是基于深度学习算法，利用已知交通事故现场信息和所述已知交通事故现场信息对应的已知交通事故责任认定信息训练得到的；生成模块，用于基于所述指定事故分析模型，根据所述交通事故现场信息，生成交通事故责任认定信息。
- 根据权利要求6所述的系统，其特征在于，所述系统还包括：请求接收模块，用于接收来自客户端的下载所述交通事故责任认定信息的下载请求；获取模块，用于响应于所述下载请求，从所述指定存储位置获取所述交通事故责任认定信息；第一发送模块，用于向所述客户端发送所述交通事故责任认定信息。
- 根据权利要求6所述的系统，其特征在于，所述系统还包括：分析模块，用于响应于接收到所述交通事故现场信息，对所述交通事故现场信息进行分析，以确定对应的指定机构；报案信息生成模块，用于生成报案信息；第三发送模块，用于将所述报案信息发送给所述指定机构。
- 根据权利要求6所述的系统，其特征在于，所述系统还包括：判断模块，用于根据所述交通事故现场信息，判断是否需要救援；第四发送模块，用于若需要救援，向指定救援机构发送救援请求信息。
- 根据权利要求9所述的系统，其特征在于，所述交通事故现场信息包括交通事故现场图像；所述判断模块具体用于：从所述交通事故现场图像中提取血迹图像，得到提取结果；根据所述提取结果，判断交通事故现场是否有血迹；若所述交通事故现场有血迹，判定需要救援；或者，若所述交通事故现场没有血迹，判定不需要救援。
- 一种服务器，其特征在于，所述服务器包括：处理器；用于存储所述处理器可执行指令的存储器；所述处理器被配置为：接收交通事故现场信息；获取指定事故分析模型，所述指定事故分析模型是基于深度学习算法，利用已知交通事故现场信息和所述已知交通事故现场信息对应的已知交通事故责任认定信息训练得到的；基于所述指定事故分析模型，根据所述交通事故现场信息，生成交通事故责任认定信息。
- 根据权利要求11所述的服务器，其特征在于，所述处理器还被配置为：接收来自客户端的下载所述交通事故责任认定信息的 下载请求；响应于所述下载请求，从所述指定存储位置获取所述交通事故责任认定信息；向所述客户端发送所述交通事故责任认定信息。
- 根据权利要求11所述的服务器，其特征在于，所述处理器还被配置为：响应于接收到所述交通事故现场信息，对所述交通事故现场信息进行分析，以确定对应的指定机构；生成报案信息；将所述报案信息发送给所述指定机构。
- 根据权利要求11所述的服务器，其特征在于，所述处理器还被配置为：根据所述交通事故现场信息，判断是否需要救援；若需要救援，向指定救援机构发送救援请求信息。
- 根据权利要求14所述的服务器，其特征在于，所述交通事故现场信息包括交通事故现场图像；所述处理器还被配置为：从所述交通事故现场图像中提取血迹图像，得到提取结果；根据所述提取结果，判断交通事故现场是否有血迹；若所述交通事故现场有血迹，判定需要救援；或者，若所述交通事故现场没有血迹，判定不需要救援。
- 一种计算机非易失性可读存储介质，所述计算机非易失性可读存储介质存储有计算机程序，其特征在于，所述计算机程序被处理器执行时实现如下步骤：接收交通事故现场信息；获取指定事故分析模型，所述指定事故分析模型是基于深度学习算法，利用已知交通事故现场信息和所述已知交通事故现场信息对应的已知交通事故责任认定信息训练得到的；基于所述指定事故分析模型，根据所述交通事故现场信息，生成交通事故责任认定信息。
- 根据权利要求16所述的计算机非易失性可读存储介质，其特征在于，所述计算机程序被处理器执行时还实现如下步骤：接收来自客户端的下载所述交通事故责任认定信息的下载请求；响应于所述下载请求，从所述指定存储位置获取所述交通事故责任认定信息；向所述客户端发送所述交通事故责任认定信息。
- 根据权利要求16所述的计算机非易失性可读存储介质，其特征在于，所述计算机程序被处理器执行时还实现如下步骤：响应于接收到所述交通事故现场信息，对所述交通事故现场信息进行分析，以确定对应的指定机构；生成报案信息；将所述报案信息发送给所述指定机构。
- 根据权利要求16所述的计算机非易失性可读存储介质，其特征在于，所述计算机程序被处理器执行时还实现如下步骤：根据所述交通事故现场信息，判断是否需要救援；若需要救援，向指定救援机构发送救援请求信息。
- 根据权利要求19所述的计算机非易失性可读存储介质，其特征在于，所述交通事故现场信息包括交通事故现场图像；所述计算机程序被处理器执行时还实现如下步骤：从所述交通事故现场图像中提取血迹图像，得到提取结果；根据所述提取结果，判断交通事故现场是否有血迹；若所述交通事故现场有血迹，判定需要救援；或者，若所述交通事故现场没有血迹，判定不需要救援。
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810690318.5 | 2018-06-28 | ||
| CN201810690318.5A CN109242739A (zh) | 2018-06-28 | 2018-06-28 | 交通事故处理方法、系统及服务器 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2020000834A1 true WO2020000834A1 (zh) | 2020-01-02 |
Family
ID=65072111
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2018/113400 Ceased WO2020000834A1 (zh) | 2018-06-28 | 2018-11-01 | 交通事故处理方法、系统及服务器 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN109242739A (zh) |
| WO (1) | WO2020000834A1 (zh) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114330087A (zh) * | 2020-09-29 | 2022-04-12 | 宝马股份公司 | 交通事故管理方法和交通事故管理系统 |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110070642A (zh) * | 2019-03-22 | 2019-07-30 | 天津大学 | 一种基于深度学习的交通事故责任评估方法及装置 |
| CN109961056A (zh) * | 2019-04-02 | 2019-07-02 | 浙江科技学院 | 基于决策树算法的交通事故责任认定方法、系统及设备 |
| CN112712691A (zh) * | 2019-10-24 | 2021-04-27 | 广州汽车集团股份有限公司 | 一种智慧交通事故处理方法及装置 |
| CN111161533B (zh) * | 2019-12-04 | 2021-11-02 | 支付宝(杭州)信息技术有限公司 | 交通事故处理方法和装置、电子设备 |
| CN112233421A (zh) * | 2020-10-15 | 2021-01-15 | 胡歆柯 | 一种基于机器视觉的城市智慧交通监控智能系统 |
| CN112381379A (zh) * | 2020-11-11 | 2021-02-19 | 深圳市元征科技股份有限公司 | 一种交通事故的处理方法、装置及电子设备 |
| CN115394099B (zh) * | 2022-08-18 | 2024-02-02 | 杭州亥迪交通设备有限公司 | 道路交通诱导方法、系统、设备及交通指引牌 |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106101628A (zh) * | 2016-06-30 | 2016-11-09 | 深圳市元征科技股份有限公司 | 一种车辆事故处理的方法以及终端 |
| CN106504173A (zh) * | 2016-12-19 | 2017-03-15 | 东软集团股份有限公司 | 交通事故处理的方法、装置及系统 |
| CN106781436A (zh) * | 2016-12-19 | 2017-05-31 | 东软集团股份有限公司 | 交通事故处理方法及装置 |
| WO2017193933A1 (zh) * | 2016-05-10 | 2017-11-16 | 腾讯科技（深圳）有限公司 | 交通事故预警方法及交通事故预警装置 |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20060014765A (ko) * | 2004-08-12 | 2006-02-16 | 주식회사 현대오토넷 | 텔레매틱스 시스템을 이용한 긴급 구난 서비스 시스템 및방법 |
| CN103824154A (zh) * | 2014-02-08 | 2014-05-28 | 刘文书 | 一种车在线事故解决方法及系统 |
| CN107332875A (zh) * | 2017-05-27 | 2017-11-07 | 上海与德科技有限公司 | 一种网约车监控方法及系统、云端服务器 |
| CN107590762A (zh) * | 2017-08-24 | 2018-01-16 | 成都安程通科技有限公司 | 一种车载事故自动上报的系统 |
| CN107909113B (zh) * | 2017-11-29 | 2021-11-16 | 北京小米移动软件有限公司 | 交通事故图像处理方法、装置及存储介质 |
| CN108154696A (zh) * | 2017-12-25 | 2018-06-12 | 重庆冀繁科技发展有限公司 | 车辆事故管理系统及方法 |
-
2018
- 2018-06-28 CN CN201810690318.5A patent/CN109242739A/zh active Pending
- 2018-11-01 WO PCT/CN2018/113400 patent/WO2020000834A1/zh not_active Ceased
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017193933A1 (zh) * | 2016-05-10 | 2017-11-16 | 腾讯科技（深圳）有限公司 | 交通事故预警方法及交通事故预警装置 |
| CN106101628A (zh) * | 2016-06-30 | 2016-11-09 | 深圳市元征科技股份有限公司 | 一种车辆事故处理的方法以及终端 |
| CN106504173A (zh) * | 2016-12-19 | 2017-03-15 | 东软集团股份有限公司 | 交通事故处理的方法、装置及系统 |
| CN106781436A (zh) * | 2016-12-19 | 2017-05-31 | 东软集团股份有限公司 | 交通事故处理方法及装置 |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114330087A (zh) * | 2020-09-29 | 2022-04-12 | 宝马股份公司 | 交通事故管理方法和交通事故管理系统 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109242739A (zh) | 2019-01-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2020000834A1 (zh) | 交通事故处理方法、系统及服务器 | |
| US9953517B2 (en) | Risk early warning method and apparatus | |
| CN106781436B (zh) | 交通事故处理方法及装置 | |
| US9141762B2 (en) | System and method for analyzing and controlling epidemics | |
| US9843611B2 (en) | Incident data collection for public protection agencies | |
| CN114549221B (zh) | 车辆事故损失处理方法、装置、计算机设备及存储介质 | |
| CN103093614A (zh) | 基于移动终端处理交通事故的方法及装置 | |
| WO2020077563A1 (zh) | 数据存储方法、装置、电子设备及计算机可读存储介质 | |
| CN110443312B (zh) | 一种基于人体姿态的城管方法及系统 | |
| US20250071205A1 (en) | Prioritizing emergency calls based on caller response to automated query | |
| CN114550951B (zh) | 一种快速就医处理方法、装置、计算机设备及存储介质 | |
| KR100986622B1 (ko) | Ｌｂｓ 기반의 모바일 단말을 이용한 사고 민원 처리 시스템 및 방법 | |
| CN109767372B (zh) | 利用车辆运行数据进行交通事故快速定责的系统及方法 | |
| WO2021159634A1 (zh) | 基于分类回归双域模型的流量识别方法和系统 | |
| CN108495096B (zh) | 一种移动执法系统及方法 | |
| CN107590997A (zh) | 一种基于物联网的交通监控方法及装置 | |
| JP2007065854A (ja) | 道路状況報告用携帯端末、プログラム、緊急報告支援システム及び緊急報告支援方法 | |
| JP2022182966A (ja) | 事故分析装置、事故分析方法、およびプログラム | |
| CN105050037A (zh) | 一种获取等待时长的方法和装置 | |
| CN106850391B (zh) | 基于互联网的报警处置方法和系统 | |
| US20170358948A1 (en) | Systems and methods for expediting repairs of utility equipment using electronic dialogs with people | |
| CN118658297A (zh) | 智能化交通事故识别装置及方法 | |
| AU2013201326B2 (en) | A Method and System for Generating a Report | |
| US11995733B2 (en) | Method and system for linking unsolicited electronic tips to public-safety data | |
| CN113537923A (zh) | 一种建筑信息处理方法、物联信息控制器及智慧城市系统 |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 18924079 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 18924079 Country of ref document: EP Kind code of ref document: A1 |