CN113220880B - Traffic incident evaluation method, system, terminal and medium based on semantic analysis - Google Patents

Abstract

The invention discloses a traffic event evaluation method, system, terminal and medium based on semantic analysis, and relates to the technical field of intelligent transportation. Analyze and obtain the evaluation results of traffic incidents from the current subjective angle; judge whether the evaluation results of traffic incidents from different subjective angles are consistent: if they are consistent, select one of the evaluation results of traffic incidents as the final evaluation result; if they are inconsistent, use all subjective information Or the re-acquired accident image information establishes a traffic event evaluation standard, and the traffic event evaluation result with the highest degree of acquaintance with the traffic event evaluation standard is used as the final evaluation result. The invention only needs to provide corresponding language information to complete responsibility determination after a traffic incident occurs, without waiting for relevant personnel to arrive at the scene for processing, and at the same time overcomes the limitation of image acquisition.

Description

Traffic incident evaluation method, system, terminal and medium based on semantic analysis

technical field

The invention relates to the technical field of intelligent transportation, and more particularly, to a method, system, terminal and medium for judging traffic events based on semantic analysis.

Background technique

Traffic incidents generally refer to incidents in which a vehicle causes personal injury or property damage due to fault or accident on the road. The responsibility determination of existing traffic incidents generally requires the traffic police or insurance company professionals to make judgments. In this process, waiting for a long time for the traffic police or professionals is likely to cause road congestion, which is not conducive to the normal operation of urban roads, and is likely to cause secondary traffic. Therefore, the realization of intelligent evaluation of traffic incidents is a key research direction in the field of traffic technology at present.

At present, the intelligent evaluation of traffic events is mainly based on the analysis and processing of image information or video information collected by the image acquisition system of the vehicle or road layout. However, the accident vehicles in traffic incidents have no equipment such as driving recorders installed. Even if the driving recorders are installed, there is a certain acquisition angle. In addition, due to the limitation of the project scale, not all road sections are equipped with cameras. As a result, there are certain limitations in the intelligent evaluation of traffic incidents based on image information or video information.

Therefore, how to study and design a traffic incident evaluation method, system, terminal and medium based on semantic analysis is an urgent problem we need to solve at present.

SUMMARY OF THE INVENTION

In order to solve the deficiencies in the prior art, the purpose of the present invention is to provide a traffic event evaluation method, system, terminal and medium based on semantic analysis.

The above-mentioned technical purpose of the present invention is achieved through the following technical solutions:

In the first aspect, a traffic incident evaluation method based on semantic analysis is provided, which includes the following steps:

Obtain the voice information of each event party in the traffic event describing the traffic event;

Subjective information is obtained after feature extraction of speech information by semantic analysis method, and road distribution information, intentional driving information and event occurrence information under the corresponding subjective angle are obtained after feature classification of subjective information;

According to the simulation analysis of road distribution information, intentional driving information, and event occurrence information, the traffic incident judgment results under the current subjective perspective are obtained, and the traffic incident judgment results under different subjective perspectives are obtained after processing the voice information of each incident party according to the above steps;

Determine whether the traffic incident evaluation results from different subjective angles are consistent:

If they are consistent, select one of the traffic incident evaluation results as the final evaluation result;

If they are inconsistent, establish a traffic event evaluation standard based on all the subjective information or the re-acquired accident image information, and take the traffic event evaluation result with the highest degree of acquaintance with the traffic event evaluation standard as the final evaluation result.

Further, the analysis process of the traffic incident evaluation result is specifically:

Establish a road plan according to the road distribution information, and determine the trajectory points of the traffic event in the road plan according to the event occurrence information, and the trajectory points include at least a start point and an end point;

According to the intentional driving information and the accident type in the event occurrence information, the simulated driving trajectory of each event party in the road plan under the current subjective perspective is obtained by simulation analysis respectively;

By simulating the synchronous movement of each event party according to the corresponding simulated driving trajectory, and analyzing the results of the synchronous movement, the traffic incident evaluation results from the current subjective perspective are obtained, and the voice information of each event party is processed according to the above steps to obtain traffic incidents from different subjective perspectives. Traffic incident evaluation results.

Further, the specific process of establishing traffic incident evaluation criteria based on all subjective information is as follows:

Obtain the same temporal features in all subjective information as basic event information;

Establish a standard driving trajectory framework according to the basic event information, and use the event logic rationality method to refer to the basic event information analysis to obtain at least one set of missing supplementary feature combinations;

After perfecting the standard driving trajectory framework according to the combination of supplementary features, at least one standard simulated trajectory is obtained, and all standard simulated trajectories are used as the evaluation criteria for traffic incidents for final evaluation.

Further, the specific process of establishing traffic incident evaluation criteria according to the re-acquired accident image information is as follows:

The intuitive event information in the accident image information is extracted by the image recognition method. The intuitive event information includes the starting point, the ending point, the vehicle distribution information, the tire steering information, the tire trace information, and the road plane information;

At least one virtual driving trajectory is constructed according to the intuitive event information, and at least one standard virtual driving trajectory is obtained by screening from the virtual driving trajectory according to all subjective information as a traffic event evaluation criterion for final evaluation.

Further, the calculation process of the degree of acquaintance between the traffic incident judgment result and the traffic incident judgment standard is specifically:

After processing the information according to the traffic incident judging standard, the standard judging result is obtained;

Whether the judgment result of the judgment standard is the same as the judgment result of the traffic incident;

If they are the same, the same principle is used to calculate the similarity value under the current traffic incident evaluation criteria; the formula for calculating the similarity value according to the same principle is as follows:

Among them, S _T represents the similarity value under the same principle; N represents the total number of features in the evaluation result of the traffic incident; N _T represents the number of features in the evaluation result of the traffic event that are the same as the evaluation standard of the corresponding traffic event;

If they are not the same, use the opposite principle to calculate the similarity value under the current traffic incident evaluation criteria; the formula for calculating the similarity value according to the opposite principle is as follows:

Among them, S _T represents the similarity value under the opposite principle;

According to the sum of all similarity values, the similarity of the evaluation results of the corresponding traffic events is obtained, and the similarity of the evaluation results of different traffic events is compared to determine the final evaluation result.

Further, the acquisition process of the voice information is specifically:

The corresponding voice description template is called according to the accident type, and the voice description template is arranged by multiple sub-classification modules according to the accident logic;

Each sub-category module matches and uploads associated professional terms from the database for display;

Voice information is recorded when the sub-classification module is turned on, and classified and saved when it is turned off.

Further, the starting process of the traffic incident judgment is specifically:

Obtain the judgment request information, which at least includes the identity information of the accident party, the accident type, and the accident location;

Call the voice description template according to the accident type, and combine the identity information to generate a description link for each accident party to independently access the voice description template;

The description link is sent to the user terminal of the corresponding accident party according to the identity information.

In the second aspect, a traffic incident evaluation system based on semantic analysis is provided, including:

The voice acquisition module is used to acquire the voice information of each event party in the traffic event describing the traffic event;

The semantic analysis module is used to obtain subjective information after feature extraction of speech information through semantic analysis method, and to obtain road distribution information, intentional driving information, and event occurrence information under the corresponding subjective angle after feature classification of the subjective information;

The first evaluation module is used to obtain the evaluation results of traffic incidents from the current subjective perspective according to the road distribution information, the intentional driving information, and the event occurrence information, and to obtain the traffic incident evaluation results from different subjective perspectives after processing the voice information of each event party according to the above steps. The result of the traffic incident evaluation;

The second evaluation module is used to determine whether the evaluation results of traffic incidents from different subjective angles are consistent: if they are consistent, select one of the traffic event evaluation results as the final evaluation result; The image information establishes a traffic event evaluation standard, and the traffic event evaluation result with the highest degree of acquaintance with the traffic event evaluation standard is used as the final evaluation result.

In a third aspect, a computer terminal is provided, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, when the processor executes the program, any one of the first aspect can be implemented. The described traffic incident evaluation method based on semantic analysis.

In a fourth aspect, there is provided a computer-readable medium on which a computer program is stored, the computer program being executed by a processor to implement the method for judging traffic events based on semantic analysis according to any one of the first aspects.

Compared with the prior art, the present invention has the following beneficial effects:

1. The present invention recognizes the feature information in the speech information through the semantic analysis method, restores the driving trajectory of the event occurrence process through the feature information, and obtains the final judgment result through the driving trajectory analysis, and only needs to provide the corresponding language information after the traffic incident occurs. Responsibility determination can be completed without waiting for relevant personnel to arrive at the scene for processing, and it also overcomes the limitation of image collection;

2. When there is an objection to the final judgment result due to the difference in the description of the accident parties, the present invention can also establish a standard simulated trajectory after data processing based on all subjective information, and use the standard simulated trajectory as the traffic incident judgment standard. The results can effectively eliminate the differences caused by the phonetic description;

3. In the present invention, when there is an error in the final judgment result because the accident party deliberately hides and changes the facts, the image processing technology can be used to process the accident image information uploaded by the accident party to establish a standard virtual driving trajectory, which can effectively offset the subjective factors. error effects;

4. The present invention compares and analyzes the traffic incident judging standard of the primary judgment and the traffic incident judging standard obtained by the secondary processing through the same principle and the opposite principle to obtain the final judgment result, which can effectively make up for the processing result caused by the lack of data. questions of accuracy;

5. The whole evaluation process of the present invention is simple and convenient to operate, benefits from the popularization of intelligent equipment, does not need to modify the accident vehicle and road supporting facilities, and has low cost of promotion and application.

Description of drawings

The accompanying drawings described herein are used to provide further understanding of the embodiments of the present invention, and constitute a part of the present application, and do not constitute limitations to the embodiments of the present invention. In the attached image:

Fig. 1 is the flow chart in the embodiment of the present invention;

FIG. 2 is a system block diagram in an embodiment of the present invention.

Detailed ways

In order to make the purpose, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and the accompanying drawings. as a limitation of the present invention.

Embodiment 1: A traffic event evaluation method based on semantic analysis, as shown in FIG. 1 , is specifically implemented by the following steps.

S1: Acquire the voice information of each event party in the traffic event describing the traffic event.

The acquisition process of the voice information is as follows: according to the accident type, the corresponding voice description template is called, and the voice description templates are arranged by multiple sub-classification modules according to the accident logic; each sub-classification module matches and uploads the associated professional terms from the database for display; Voice information is recorded when the sub-classification module is turned on, and classified and saved when it is turned off.

It should be noted that the sub-classification module includes but is not limited to road type, intended driving type and accident type. The accident party can describe the voice information provided by the sub-classification module according to the description logic and technical terms in the sub-classification module, which can make the entire description process sequence. , reducing the difficulty of semantic analysis, and effectively reducing the occurrence of inaccurate speech information caused by the accident party intentionally concealing real information or differences in expression ability.

The start-up process of traffic incident evaluation is as follows: obtain evaluation request information, which includes at least the identity information of the accident party, accident type, and accident location; call the voice description template according to the accident type, and combine the identity information to generate independent access to each accident party. The description link of the voice description template; the description link is sent to the user terminal of the corresponding accident party according to the identity information.

It should be noted that the judgment request information can be provided by any one of the accident parties, or provided by multiple accident parties at the same time. When there are different judgment request information, it can be processed for record, and can also be confirmed by feedback.

S2: Subjective information is obtained after feature extraction of speech information by semantic analysis method, and road distribution information, intentional driving information, and event occurrence information under the corresponding subjective angle are obtained after feature classification of subjective information. It should be noted that the subjective perspective is the accident party who provides the voice information.

S3: According to the simulation analysis of road distribution information, intentional driving information, and event occurrence information, the evaluation results of traffic incidents from the current subjective perspective are obtained, and the evaluation results of traffic incidents from different subjective perspectives are obtained after processing the voice information of each event party according to the above steps. .

The analysis process of the traffic incident evaluation results is as follows: establishing a road plan according to the road distribution information, and determining the trajectory points of the traffic incident in the road plan according to the event occurrence information. Trajectory points, such as crossing nodes, braking nodes, acceleration nodes, turning nodes, etc. According to the intentional driving information and the accident type in the event occurrence information, the simulated driving trajectory of each event party in the road plan under the current subjective perspective is obtained by simulation analysis. By simulating the synchronous movement of each event party according to the corresponding simulated driving trajectory, and analyzing the results of the synchronous movement, the traffic incident evaluation results from the current subjective perspective are obtained, and the voice information of each event party is processed according to the above steps to obtain traffic incidents from different subjective perspectives. Traffic incident evaluation results. Determine whether the evaluation results of traffic incidents from different subjective angles are consistent: if they are consistent, select one of the traffic event evaluation results as the final evaluation result; if they are inconsistent, establish traffic event evaluation criteria based on all subjective information or re-acquired accident image information , and take the traffic incident judgment result with the highest degree of acquaintance with the traffic incident judgment standard as the final judgment result.

The specific process of establishing a traffic event evaluation standard based on all subjective information is as follows: obtaining the same time characteristics in all subjective information as the basic event information; establishing a standard driving trajectory frame based on the basic event information, and referring to the basic event information by means of event logic and rationality At least one set of missing supplementary feature combinations is obtained by analysis; after the standard driving trajectory framework is perfected according to the supplementary feature combinations, at least one standard simulated trajectory is obtained, and all the standard simulated trajectories are used as the traffic incident evaluation criteria for final evaluation.

In addition, the specific process of establishing traffic incident evaluation criteria based on the re-acquired accident image information is as follows: extracting the intuitive event information in the accident image information by image recognition method, the intuitive event information includes the starting point, ending point, vehicle distribution information, Tire steering information, tire trace information, road plane information; construct at least one virtual driving trajectory based on the intuitive event information, and screen out at least one standard virtual driving trajectory from the virtual driving trajectory based on all subjective information as the traffic event evaluation criteria for final evaluation .

The calculation process of the degree of acquaintance between the evaluation result of the traffic incident and the evaluation standard of the traffic event is as follows: the standard evaluation result is obtained after information processing according to the evaluation standard of the traffic event; whether the evaluation result of the evaluation standard is the same as the evaluation result of the traffic event.

If they are the same, the same principle is used to calculate the similarity value under the current traffic incident evaluation criteria; the formula for calculating the similarity value according to the same principle is as follows:

Among them, S _T represents the similarity value under the same principle; N represents the total number of features in the evaluation result of the traffic incident; N _T represents the number of features in the evaluation result of the traffic event that are the same as the evaluation standard of the corresponding traffic event.

If they are not the same, use the opposite principle to calculate the similarity value under the current traffic incident evaluation criteria; the formula for calculating the similarity value according to the opposite principle is as follows:

Among them, S _T represents the similarity value under the opposite principle;

According to the sum of all similarity values, the similarity of the evaluation results of the corresponding traffic events is obtained, and the similarity of the evaluation results of different traffic events is compared to determine the final evaluation result.

Embodiment 2: a traffic event evaluation system based on semantic analysis, including a speech acquisition module, a semantic analysis module, a first evaluation module, and a second evaluation module. The voice acquisition module is used to acquire the voice information of each event party in the traffic event describing the traffic event. The semantic analysis module is used to obtain subjective information after feature extraction of speech information through semantic analysis method, and to obtain road distribution information, intentional driving information, and event occurrence information from a corresponding subjective perspective after feature classification of the subjective information. The first evaluation module is used to obtain the traffic incident evaluation results from the current subjective perspective according to the road distribution information, the intentional driving information, and the event occurrence information, and to obtain the traffic incident evaluation results from different subjective perspectives after processing the voice information of each event party according to the above steps. The results of traffic incident evaluation. The second evaluation module is used to determine whether the evaluation results of traffic incidents from different subjective angles are consistent: if they are consistent, select one of the traffic event evaluation results as the final evaluation result; The image information establishes a traffic event evaluation standard, and the traffic event evaluation result with the highest degree of acquaintance with the traffic event evaluation standard is used as the final evaluation result.

As will be appreciated by those skilled in the art, the embodiments of the present application may be provided as a method, a system, or a computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It will be understood that each flow and/or block in the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to the processor of a general purpose computer, special purpose computer, embedded processor or other programmable data processing device to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing device produce Means for implementing the functions specified in a flow or flow of a flowchart and/or a block or blocks of a block diagram.

These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory result in an article of manufacture comprising instruction means, the instructions The apparatus implements the functions specified in the flow or flow of the flowcharts and/or the block or blocks of the block diagrams.

These computer program instructions can also be loaded on a computer or other programmable data processing device to cause a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process such that The instructions provide steps for implementing the functions specified in the flow or blocks of the flowcharts and/or the block or blocks of the block diagrams.

The above specific embodiments further describe the purpose, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above are only specific embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Within the spirit and principle of the present invention, any modifications, equivalent replacements, improvements, etc. made should be included within the protection scope of the present invention.

Claims (9)

1. The traffic incident judgment method based on semantic analysis is characterized by comprising the following steps of:

acquiring voice information describing the traffic events by each event party in the traffic events;

carrying out feature extraction on voice information by a semantic analysis method to obtain subjective information, and carrying out feature classification on the subjective information to obtain road distribution information, intention driving information and event occurrence information under a corresponding subjective angle;

according to the road distribution information, the intention driving information and the incident occurrence information, a traffic incident judgment result under the current subjective angle is obtained through simulation analysis, and according to the steps, the voice information of each incident party is processed to obtain the traffic incident judgment results under different subjective angles;

judging whether the judgment results of the traffic events under different subjective angles are consistent or not:

if the traffic events are consistent, selecting one of the traffic event judgment results as a final judgment result;

if the judgment result is inconsistent with the judgment result, establishing a traffic incident judgment standard according to all subjective information or the reacquired accident image information, and taking a traffic incident judgment result with the highest degree of recognition with the traffic incident judgment standard as a final judgment result;

the calculation process of the recognition degree of the traffic incident judgment result and the traffic incident judgment standard specifically comprises the following steps:

processing information according to the traffic incident judgment standard to obtain a standard judgment result;

judging whether the standard judgment result is the same as the traffic event judgment result or not;

if the traffic events are the same, calculating a similarity value under the current traffic event judgment standard by adopting the same principle; the formula for calculating the similarity value by the same principle is specifically as follows:

wherein S is_TRepresenting similarity values under the same principle; n represents the total characteristic quantity in the judgment result of the traffic incident; n is a radical of_TRepresenting the characteristic quantity which is the same as the judgment standard of the corresponding traffic incident in the judgment result of the traffic incident;

if not, calculating a similarity value under the current traffic event judgment standard by adopting an opposite principle; the formula for calculating the similarity value by the principle of the inverse method is specifically as follows:

wherein S is_TRepresenting similar values under the opposite principle;

and obtaining the similarity of the judgment results of the corresponding traffic events according to the sum of all the similarity values, and comparing the similarities of the judgment results of different traffic events to determine the final judgment result.

2. The traffic incident evaluation method based on semantic analysis according to claim 1, wherein the analysis process of the traffic incident evaluation result specifically comprises:

establishing a road plane graph according to the road distribution information, and determining track points of the traffic incident in the road plane graph according to the incident occurrence information, wherein the track points at least comprise a starting point and an ending point;

respectively carrying out simulation analysis according to accident types in the intention driving information and the event occurrence information to obtain simulated driving tracks of all event parties in a road plan under the current subjective angle;

and processing the voice information of each incident party according to the steps to obtain the traffic incident judgment results under different subjective angles.

3. The traffic incident evaluation method based on semantic analysis according to claim 2, wherein the specific process of establishing the traffic incident evaluation criterion according to all subjective information comprises:

acquiring the same time characteristics in all subjective information as basic event information;

establishing a standard driving track frame according to the basic event information, and analyzing by referring to the basic event information in an event logic rationality mode to obtain at least one group of missing supplementary feature combinations;

and (4) combining and perfecting the standard driving track framework according to the supplementary features to obtain at least one standard simulation track, and taking all the standard simulation tracks as the traffic event judgment standard for final judgment.

4. The semantic analysis based traffic incident evaluation method according to claim 2, wherein the specific process of establishing the traffic incident evaluation criterion based on the retrieved accident image information is as follows:

visual event information in the accident image information is extracted through an image recognition method, wherein the visual event information comprises a starting point, an ending point, vehicle distribution information, tire steering information, tire trace information and road plane information of an event;

and constructing at least one virtual driving track according to the visual event information, and screening at least one standard virtual driving track from the virtual driving tracks according to all subjective information to be used as a traffic event judgment standard for final judgment.

5. The traffic incident assessment method based on semantic analysis according to any one of claims 1 to 4, wherein the voice information is obtained by:

calling a corresponding voice description template according to the accident type, wherein the voice description template is arranged by a plurality of sub-classification modules according to accident logic;

each sub-classification module is matched with and uploads related professional terms from the database to display;

and recording voice information when the sub-classification module is opened, and performing classification storage when the sub-classification module is closed.

6. The traffic event assessment method based on semantic analysis according to any one of claims 1 to 4, wherein the starting process of the traffic event assessment specifically comprises:

acquiring judgment request information, wherein the judgment request information at least comprises identity information of an accident party, an accident type and an accident site;

calling a voice description template according to the accident type, and generating a description link of each accident party for independently accessing the voice description template by combining identity information;

and sending the description link to the user terminal of the corresponding accident party according to the identity information.

7. The traffic incident judgment system based on semantic analysis is characterized by comprising the following components:

the voice acquisition module is used for acquiring voice information describing the traffic events of all event parties in the traffic events;

the semantic analysis module is used for extracting the characteristics of the voice information through a semantic analysis method to obtain subjective information, and classifying the characteristics of the subjective information to obtain road distribution information, intention driving information and event occurrence information under the corresponding subjective angle;

the first evaluation module is used for obtaining a traffic event evaluation result under the current subjective angle according to the road distribution information, the intention driving information and the event occurrence information through simulation analysis, and obtaining the traffic event evaluation result under different subjective angles after processing the voice information of each event party according to the first evaluation module;

the second judging module is used for judging whether the judging results of the traffic events under different subjective angles are consistent or not: if the traffic events are consistent, selecting one of the traffic event judgment results as a final judgment result; if the judgment result is inconsistent with the judgment result, establishing a traffic incident judgment standard according to all subjective information or the reacquired accident image information, and taking a traffic incident judgment result with the highest degree of recognition with the traffic incident judgment standard as a final judgment result;

the calculation process of the recognition degree of the traffic incident judgment result and the traffic incident judgment standard specifically comprises the following steps:

processing information according to the traffic incident judgment standard to obtain a standard judgment result;

judging whether the standard judgment result is the same as the traffic event judgment result or not;

if the traffic events are the same, calculating a similarity value under the current traffic event judgment standard by adopting the same principle; the formula for calculating the similarity value by the same principle is specifically as follows:

wherein S is_TRepresenting similarity values under the same principle; n represents the total characteristic quantity in the judgment result of the traffic incident; n is a radical of_TRepresenting the characteristic quantity which is the same as the judgment standard of the corresponding traffic incident in the judgment result of the traffic incident;

if not, calculating a similarity value under the current traffic event judgment standard by adopting an opposite principle; the formula for calculating the similarity value by the principle of the inverse method is specifically as follows:

wherein S is_TRepresenting similar values under the opposite principle;

and obtaining the similarity of the judgment results of the corresponding traffic events according to the sum of all the similarity values, and comparing the similarities of the judgment results of different traffic events to determine the final judgment result.

8. A computer terminal comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the traffic event judgment method based on semantic analysis according to any one of claims 1 to 6.

9. A computer-readable medium, on which a computer program is stored, the computer program being executable by a processor for implementing a traffic event assessment method based on semantic analysis according to any one of claims 1 to 6.

Images