Road traffic multi-source information collection method and device and storage medium

Abstract

This application relates to the field of intelligent transportation technology, and in particular to a method, device, and storage medium for collecting multi-source information on road traffic. The method acquires raw traffic data from at least two heterogeneous sources in a road traffic scenario, and arranges it based on source identification information, acquisition time sequence information, and spatial location description information to obtain a multi-source traffic data set. It then extracts data quality characterization information, scene state characterization information, and object association clue information corresponding to each raw traffic data, constructs dynamic confidence characterization results for each acquisition source, and performs cross-source alignment and conflict discrimination on the multi-source traffic data set to obtain a unified traffic description fragment set corresponding to the target traffic object. Based on the unified traffic description fragment set, a multi-source collaborative acquisition instruction set corresponding to the target traffic object is generated. Adaptive acquisition is performed based on the multi-source collaborative acquisition instruction set, and the unified traffic description fragment set is updated based on the returned incremental traffic data.

Description

Technical Field

[0001] This application relates to the field of intelligent transportation technology, and in particular to a method, device and storage medium for collecting multi-source information on road traffic. Background Technology

[0002] With the continuous development of smart cities and intelligent transportation systems, the demand for information perception in road traffic scenarios is constantly increasing. To effectively grasp the road's operational status, the activities of traffic participants, and sudden traffic events, existing technologies typically deploy various types of data collection devices or connect to multiple information sources in road traffic scenarios to obtain road traffic-related data. These information sources typically include video capture devices, radar devices, geomagnetic detection devices, roadside sensing units, vehicle-mounted terminals, and other traffic information collection devices or platforms. Because different data sources differ in their collection mechanisms, data structures, spatiotemporal granularity, and information expression methods, road traffic information usually exhibits a multi-source heterogeneous characteristic. In related technologies, the collection and processing of road traffic data typically focuses on the separate acquisition and independent use of data output from each data source, or simply the aggregation and surface stitching of multi-source data. While this processing method can improve the coverage of traffic information acquisition to some extent, the lack of an effective correlation and organization mechanism between different data collection sources makes it difficult to coordinate data temporally, correspond spatially, and unify semantically. This can easily lead to fragmented descriptions of the same traffic object across different data collection sources, making it difficult to form a consistent and complete traffic information representation. Therefore, how to improve cross-source collaboration and descriptive consistency in the process of multi-source road traffic information collection has become an urgent problem to be solved. Summary of the Invention

[0003] The main objective of this application is to provide a method, device, and storage medium for collecting multi-source information on road traffic, aiming to solve the technical problem of how to improve cross-source coordination and description consistency in the process of collecting multi-source information on road traffic.

[0004] To achieve the above objectives, this application proposes a method for collecting multi-source road traffic information, the method comprising: The system acquires raw traffic data from at least two heterogeneous data sources in a road traffic scenario, and performs association and arrangement processing on the raw traffic data based on the source identification information, acquisition time sequence information and spatial location description information corresponding to each data source to obtain a multi-source traffic data set. Based on the multi-source traffic data set, data quality characterization information, scene state characterization information, and object association clue information corresponding to each original traffic data are extracted. Based on the data quality characterization information, scene state characterization information, and object association clue information, dynamic confidence characterization results corresponding to each data collection source are constructed. Based on the dynamic confidence representation results, cross-source alignment and conflict discrimination processing are performed on the multi-source traffic data set to obtain a unified traffic description fragment set corresponding to the target traffic object, wherein the target traffic object includes at least one of road infrastructure, traffic participants and traffic events. Based on the unified traffic description fragment set, the state transmission relationship and semantic completion relationship of the target traffic object among different collection sources are established, and a multi-source collaborative collection instruction set corresponding to the target traffic object is generated. Based on the multi-source collaborative acquisition instruction set, the corresponding acquisition source is controlled to perform adaptive acquisition processing on the target traffic object, and the incremental traffic data returned after adaptive acquisition processing is injected back to update the unified traffic description fragment set.

[0005] In one embodiment, the step of acquiring raw traffic data from at least two heterogeneous acquisition sources in a road traffic scenario, and performing correlation and arrangement processing on the raw traffic data based on the source identification information, acquisition time sequence information, and spatial location description information corresponding to each acquisition source to obtain a multi-source traffic data set, includes: Acquire raw traffic data output from various heterogeneous data acquisition sources in a road traffic scenario, and determine the data acquisition source category and data arrival status corresponding to each raw traffic data. Based on the source category and arrival status of each original traffic data, the source identification information, collection time sequence information and spatial location description information corresponding to each original traffic data are determined, and the association between each original traffic data and the source identification information, the collection time sequence information and the spatial location description information is established. Based on the aforementioned relationships, the original traffic data are processed through cross-source aggregation, sequential arrangement, and location mapping to obtain a multi-source traffic data set.

[0006] In one embodiment, the step of extracting data quality characterization information, scene state characterization information, and object association clue information corresponding to each original traffic data based on the multi-source traffic data set, and constructing a dynamic confidence characterization result corresponding to each data collection source based on the data quality characterization information, the scene state characterization information, and the object association clue information, includes: Based on the multi-source traffic data set, each original traffic data is parsed and processed to determine the data integrity information, data consistency information, and data stability information corresponding to each original traffic data, so as to generate data quality characterization information corresponding to each original traffic data. Based on the multi-source traffic data set, scene element identification and object relationship identification are performed on each original traffic data to determine the traffic flow state information, road environment state information and object correspondence information corresponding to each original traffic data, so as to generate scene state representation information and object association clue information corresponding to each original traffic data. Based on the data quality characterization information, scene state characterization information, and object association clue information corresponding to each original traffic data, the data credibility and association support of each collection source are jointly characterized, and a dynamic confidence characterization result corresponding to each collection source is constructed.

[0007] In one embodiment, the step of jointly characterizing the data credibility and correlation support of each collection source based on the data quality characterization information, the scene state characterization information, and the object association clue information corresponding to each original traffic data, and constructing a dynamic confidence characterization result corresponding to each collection source, includes: Based on the data quality characterization information corresponding to each original traffic data, the data reliability evaluation information corresponding to each collection source in the current collection cycle is determined, wherein the data reliability evaluation information is used to characterize the usability and fluctuation status of the output data of each collection source. Based on the scene state representation information and object association clue information corresponding to each original traffic data, the association support evaluation information of each collection source for the target traffic object and the target traffic scene is determined. The association support evaluation information is used to characterize the degree of support of each collection source for the traffic object identification result and the scene state determination result. Based on the data credibility evaluation information and the correlation support evaluation information corresponding to each collection source, the data credibility and correlation support of each collection source are jointly characterized to construct a dynamic confidence representation result corresponding to each collection source.

[0008] In one embodiment, the step of performing cross-source alignment and conflict discrimination processing on the multi-source traffic data set based on the dynamic confidence representation results to obtain a unified traffic description fragment set corresponding to the target traffic object includes: Based on the dynamic confidence representation results, the confidence participation relationship and alignment constraint relationship corresponding to each original traffic data in the multi-source traffic data set are determined, and candidate traffic data to participate in cross-source fusion are screened according to the confidence participation relationship and the alignment constraint relationship. Based on the candidate traffic data participating in cross-source fusion, the object identification information, status description information and location association information pointing to the same target traffic object in different collection sources are matched and aligned across sources to obtain the alignment description results corresponding to each target traffic object. Based on the alignment description results corresponding to each target traffic object, and combined with the description differences between different collection sources, conflicting description content is processed for conflict discrimination, and the valid description content after discrimination is merged and organized to obtain a unified traffic description fragment set corresponding to the target traffic object.

[0009] In one embodiment, the step of establishing the state transfer relationship and semantic completion relationship of the target traffic object among different acquisition sources based on the unified traffic description fragment set, and generating a multi-source collaborative acquisition instruction set corresponding to the target traffic object, includes: Based on the unified traffic description fragment set, extract the object state description information, state change clue information and cross-source semantic association information corresponding to each target traffic object, and establish the state correspondence relationship between each target traffic object and different collection sources. Based on the state correspondence of each target traffic object and the cross-source semantic association information, the state transmission relationship and semantic completion relationship of each target traffic object between different collection sources are determined, and collaborative collection requirement information corresponding to each target traffic object is generated. Based on the collaborative collection requirement information corresponding to each target traffic object, the collaborative collection method, collection content and collection timing corresponding to different collection sources are determined, and a set of multi-source collaborative collection instructions corresponding to the target traffic object is generated.

[0010] In one embodiment, the step of controlling the corresponding acquisition source to perform adaptive acquisition processing on the target traffic object based on the multi-source collaborative acquisition instruction set, and injecting back the incremental traffic data returned after adaptive acquisition processing to update the unified traffic description fragment set, includes: Based on the multi-source collaborative acquisition instruction set, the target acquisition source, target acquisition content, and target acquisition timing corresponding to each target traffic object are determined, and the corresponding acquisition source is controlled to enter the adaptive acquisition state corresponding to the target traffic object. Based on the adaptive acquisition state corresponding to each acquisition source, incremental traffic data corresponding to the target traffic object is obtained, and based on the object correspondence and description association between the incremental traffic data and the unified traffic description fragment set, the incremental update content corresponding to the target traffic object is determined. Based on the incremental update content corresponding to the target traffic object, the unified traffic description fragment set is back-injected and updated to obtain the updated unified traffic description fragment set.

[0011] In one embodiment, the step of determining the target acquisition source, target acquisition content, and target acquisition timing corresponding to each target traffic object based on the multi-source collaborative acquisition instruction set, and controlling the corresponding acquisition source to enter the adaptive acquisition state corresponding to the target traffic object, includes: Based on the multi-source collaborative acquisition instruction set, the acquisition collaboration requirements corresponding to each target traffic object are analyzed, and the candidate acquisition source set, candidate acquisition content set, and candidate acquisition time sequence set corresponding to each target traffic object are determined. Based on the candidate collection source set, the candidate collection content set, and the candidate collection time sequence set corresponding to each target traffic object, corresponding matching and constraint filtering processes are performed to determine the target collection source, target collection content, and target collection time sequence corresponding to each target traffic object. Based on the target acquisition source, the target acquisition content, and the target acquisition timing corresponding to each target traffic object, acquisition control information is sent to the corresponding acquisition source, and the corresponding acquisition source is controlled to enter the adaptive acquisition state corresponding to the target traffic object.

[0012] Furthermore, to achieve the above objectives, this application also proposes a road traffic multi-source information collection device, which includes: The data acquisition module is used to acquire raw traffic data from at least two heterogeneous collection sources in a road traffic scenario, and to perform association and arrangement processing on the raw traffic data based on the source identification information, collection time sequence information and spatial location description information corresponding to each collection source to obtain a multi-source traffic data set. The confidence characterization module is used to extract data quality characterization information, scene state characterization information, and object association clue information corresponding to each original traffic data based on the multi-source traffic data set, and to construct dynamic confidence characterization results corresponding to each collection source based on the data quality characterization information, the scene state characterization information, and the object association clue information. The unified description module is used to perform cross-source alignment and conflict discrimination processing on the multi-source traffic data set based on the dynamic confidence representation results, so as to obtain a unified traffic description fragment set corresponding to the target traffic object, wherein the target traffic object includes at least one of road infrastructure, traffic participants and traffic events. The multi-source collaboration module is used to establish the state transmission relationship and semantic completion relationship of the target traffic object among different acquisition sources based on the unified traffic description fragment set, and generate a multi-source collaborative acquisition instruction set corresponding to the target traffic object. The back-injection update module is used to control the corresponding acquisition source to perform adaptive acquisition processing on the target traffic object based on the multi-source collaborative acquisition instruction set, and to back-inject and update the incremental traffic data returned after adaptive acquisition processing, so as to update the unified traffic description fragment set.

[0013] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the road traffic multi-source information collection method described above.

[0014] This application acquires raw traffic data from at least two heterogeneous data sources in a road traffic scenario. Based on the source identification information, acquisition time sequence information, and spatial location description information corresponding to each data source, the raw traffic data is correlated and arranged to obtain a multi-source traffic data set. Based on the multi-source traffic data set, data quality characterization information, scene state characterization information, and object association clue information corresponding to each raw traffic data are extracted. Based on the data quality characterization information, scene state characterization information, and object association clue information, dynamic confidence characterization results corresponding to each data source are constructed. Based on the dynamic confidence characterization results, the multi-source traffic data set is then... Cross-source alignment and conflict resolution processes are used to obtain a unified traffic description fragment set corresponding to the target traffic object, where the target traffic object includes at least one of road infrastructure, traffic participants, and traffic events. Based on the unified traffic description fragment set, state transfer relationships and semantic completion relationships of the target traffic object are established among different acquisition sources, generating a multi-source collaborative acquisition instruction set corresponding to the target traffic object. Based on the multi-source collaborative acquisition instruction set, the corresponding acquisition sources are controlled to perform adaptive acquisition processing on the target traffic object, and the incremental traffic data returned after adaptive acquisition processing is back-injected to update the unified traffic description fragment set. This application performs correlation and arrangement processing on raw traffic data from at least two heterogeneous collection sources, establishing a foundation for correlation between data from different collection sources in terms of source, temporal sequence, and spatial dimensions. Furthermore, by extracting data quality characterization information, scene state characterization information, and object association clues, it constructs dynamic confidence characterization results corresponding to each collection source, providing a basis for subsequent processing of multi-source data. Based on this, it performs cross-source alignment and conflict discrimination processing on the multi-source traffic data set, forming a unified traffic description fragment set corresponding to the target traffic object. It further establishes state transmission relationships and semantic completion relationships between different collection sources for the target traffic object, generating a multi-source collaborative collection instruction set, enabling different collection sources to collaboratively collect data around the same target traffic object. Combined with the injection and updating of incremental traffic data, the unified traffic description fragment set can be continuously supplemented and corrected, thereby improving cross-source collaboration and description consistency in the process of collecting multi-source road traffic information. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating the first embodiment of the road traffic multi-source information collection method of this application; Figure 2 This is a schematic diagram of a sub-process in the second embodiment of the road traffic multi-source information collection method of this application; Figure 3 This is a schematic diagram of a sub-process in the third embodiment of the road traffic multi-source information collection method of this application; Figure 4 This is a schematic diagram of the module structure of the road traffic multi-source information collection device according to an embodiment of this application; Figure 5 This is a schematic diagram of the equipment structure of the hardware operating environment involved in the road traffic multi-source information collection method in this application embodiment. Detailed Implementation

[0016] The present invention will be further described and illustrated below with reference to specific embodiments and the accompanying drawings: It should be noted that with the continuous development of smart cities and intelligent transportation systems, the demand for information perception in road traffic scenarios is constantly increasing. To effectively grasp the road's operational status, the activities of traffic participants, and sudden traffic events, existing technologies typically deploy various types of data collection devices or access multiple information sources in road traffic scenarios to obtain road traffic-related data. These information sources typically include video acquisition devices, radar devices, geomagnetic detection devices, roadside sensing units, vehicle-mounted terminals, and other traffic information collection devices or platforms. Because different data sources differ in their acquisition mechanisms, data structures, spatiotemporal granularity, and information expression methods, road traffic information usually exhibits a multi-source heterogeneous characteristic. In related technologies, the acquisition and processing of road traffic data typically focuses on the separate acquisition and independent use of data output from each data source, or simply the aggregation and surface stitching of multi-source data. While this processing method can improve the coverage of traffic information acquisition to some extent, the lack of an effective correlation and organization mechanism between different data collection sources makes it difficult to coordinate data temporally, correspond spatially, and unify semantically. This can easily lead to fragmented descriptions of the same traffic object across different data collection sources, making it difficult to form a consistent and complete traffic information representation. Therefore, how to improve cross-source collaboration and descriptive consistency in the process of multi-source road traffic information collection has become an urgent problem to be solved.

[0017] The main solution of this application is as follows: First, acquire raw traffic data from at least two heterogeneous data sources in a road traffic scenario. Then, based on the source identification information, acquisition time sequence information, and spatial location description information corresponding to each data source, perform association and arrangement processing on the raw traffic data to obtain a multi-source traffic data set. Based on the multi-source traffic data set, extract data quality representation information, scene state representation information, and object association clue information corresponding to each raw traffic data. Based on the data quality representation information, scene state representation information, and object association clue information, construct dynamic confidence representation results corresponding to each data source. Finally, based on the dynamic confidence representation results, analyze the multi-source traffic dataset. Cross-source alignment and conflict resolution are performed to obtain a unified traffic description fragment set corresponding to the target traffic object, wherein the target traffic object includes at least one of road infrastructure, traffic participants, and traffic events. Based on the unified traffic description fragment set, state transfer relationships and semantic completion relationships of the target traffic object among different acquisition sources are established, and a multi-source collaborative acquisition instruction set corresponding to the target traffic object is generated. Based on the multi-source collaborative acquisition instruction set, the corresponding acquisition source is controlled to perform adaptive acquisition processing on the target traffic object, and the incremental traffic data returned after adaptive acquisition processing is back-injected to update the unified traffic description fragment set.

[0018] This application performs correlation and arrangement processing on raw traffic data from at least two heterogeneous collection sources, establishing a foundation for correlation between data from different collection sources in terms of source, temporal sequence, and spatial dimensions. Furthermore, by extracting data quality characterization information, scene state characterization information, and object association clues, it constructs dynamic confidence characterization results corresponding to each collection source, providing a basis for subsequent processing of multi-source data. Based on this, it performs cross-source alignment and conflict discrimination processing on the multi-source traffic data set, forming a unified traffic description fragment set corresponding to the target traffic object. It further establishes state transmission relationships and semantic completion relationships between different collection sources for the target traffic object, generating a multi-source collaborative collection instruction set, enabling different collection sources to collaboratively collect data around the same target traffic object. Combined with the injection and updating of incremental traffic data, the unified traffic description fragment set can be continuously supplemented and corrected, thereby improving cross-source collaboration and description consistency in the process of collecting multi-source road traffic information.

[0019] It should be noted that the executing entity of the method in this embodiment can be a computing service device with data processing, network communication, and program execution functions, or it can be the aforementioned road traffic multi-source information collection device with the same or similar functions. This embodiment and the following embodiments will be described using a road traffic multi-source information collection device as an example.

[0020] Based on this, a first embodiment of the road traffic multi-source information collection method of this application is proposed. Please refer to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the road traffic multi-source information collection method of this application.

[0021] In this embodiment, the method includes the following steps: S1: Acquire raw traffic data from at least two heterogeneous collection sources in a road traffic scenario, and perform association and arrangement processing on the raw traffic data based on the source identification information, collection time sequence information and spatial location description information corresponding to each collection source to obtain a multi-source traffic data set; S2: Based on the multi-source traffic data set, extract the data quality characterization information, scene state characterization information, and object association clue information corresponding to each original traffic data, and construct a dynamic confidence characterization result corresponding to each collection source based on the data quality characterization information, the scene state characterization information, and the object association clue information. It should be noted that "road traffic scenario" refers to the actual application environment related to road traffic operation. Heterogeneous acquisition sources refer to different information acquisition sources with differences in acquisition methods, sensing principles, output data types, or data structures. Raw traffic data refers to traffic-related data directly output by each acquisition source without deep fusion processing. Source identification information refers to information used to identify the acquisition source to which the raw traffic data belongs. Acquisition time sequence information refers to information characterizing the time sequence of the raw traffic data's generation, acquisition time, or time correlation. Spatial location description information refers to information characterizing the spatial region, location range, or spatial correlation corresponding to the raw traffic data. Association and arrangement processing refers to the process of organizing, collecting, and sorting raw traffic data from different acquisition sources according to predetermined association rules. Multi-source traffic data set refers to an organized data set containing traffic data from multiple acquisition sources after association and arrangement processing. Data quality characterization information refers to data description information used to reflect the availability, completeness, stability, consistency, or reliability of the raw traffic data. Scenario state characterization information refers to information used to characterize the current operating state, environmental state, or traffic state of the road traffic scenario. Object association clues refer to clues used to reflect whether different original traffic data point to the same traffic object or have a correlation. Dynamic confidence characterization results refer to results formed by comprehensively considering data quality characterization information, scene state characterization information, and object association clues, which can characterize the current data participation value and credibility of each collection source.

[0022] Specifically, the first step is to access raw traffic data from at least two heterogeneous data sources within a road traffic scenario. Since different data sources typically differ in data output format, collection cycle, and sensing location, it's necessary to first establish a basic correspondence between various types of raw traffic data based on the source identification information, collection time sequence information, and spatial location description information corresponding to each source. This can be understood as: first clearly labeling each piece of raw traffic data with "where it comes from, when it was collected, and what spatial region it corresponds to," and then collecting, associating, and arranging the raw traffic data according to this information, so that the originally scattered traffic data from various sources form a multi-source traffic data set with a unified organizational relationship. This provides a data foundation for subsequent further analysis focusing on the same scenario, the same region, or the same traffic object.

[0023] Furthermore, after obtaining the multi-source traffic data set, multiple types of representational information are extracted from each original traffic data set to describe the data's own state and its relational state with the scene. One part of this information describes the quality of the original traffic data itself, such as its completeness, stability, and consistency with similar data. Another part describes the scene state corresponding to the data, such as traffic operation status, environmental status, or object status. A third part describes whether there are object-level correlations between this data and other original traffic data. Subsequently, the data quality representation information, scene state representation information, and object correlation clue information are combined to comprehensively characterize the data credibility and participation value of each collection source in the current scene, thereby constructing a dynamic confidence representation result corresponding to each collection source. This result is not a static labeling of the collection sources, but rather dynamically adjusted as the original traffic data and scene change.

[0024] This step first correlates and organizes raw traffic data from different heterogeneous collection sources using source identification information, collection time sequence information, and spatial location description information. This establishes basic relationships between the multi-source data at the source, time, and spatial levels, preventing data from being fragmented. Based on this, it further extracts data quality characterization information, scene state characterization information, and object association clues, and constructs dynamic confidence characterization results corresponding to each collection source. This ensures that data from different collection sources can not only be organized within the same data framework but also be assigned corresponding participation criteria based on their data state, scene association, and object association. Therefore, this step provides a more targeted processing foundation for subsequent cross-source alignment, conflict detection, and unified description of multi-source traffic data.

[0025] S3: Based on the dynamic confidence representation results, cross-source alignment and conflict discrimination processing are performed on the multi-source traffic data set to obtain a unified traffic description fragment set corresponding to the target traffic object, wherein the target traffic object includes at least one of road infrastructure, traffic participants and traffic events; S4: Based on the unified traffic description fragment set, establish the state transfer relationship and semantic completion relationship of the target traffic object among different collection sources, and generate a multi-source collaborative collection instruction set corresponding to the target traffic object; It should be noted that the unified traffic description fragment set refers to a unified description result set formed by aligning, discriminating, and organizing relevant descriptions from different data collection sources for a target traffic object. State transitivity refers to the state continuation, state mapping, or state evolution relationship of the target traffic object between different data collection sources. Semantic completion refers to the complementary relationship between the descriptive information provided by different data collection sources for the same target traffic object. The multi-source collaborative data collection instruction set refers to the set of instructions generated based on the unified traffic description results, used to coordinate subsequent data collection by different data collection sources for the target traffic object.

[0026] Specifically, firstly, based on the dynamic confidence representation results, data related to the same target traffic object in the multi-source traffic dataset are filtered and organized. Since different data sources may describe the same target traffic object at different times, with different location descriptions, different granularities, or different information focuses, it is necessary to use the dynamic confidence representation results to differentiate the degree of data participation from each data source, and further perform cross-source alignment processing on the descriptions related to the target traffic object from different data sources. This alignment process can revolve around the target traffic object's identity, spatial affiliation, status content, and temporal relationships, thereby merging and corresponding the object descriptions originally scattered across different data sources. After the correspondence is completed, conflict judgment is performed on the descriptions that differ between different data sources, distinguishing which descriptions can be retained and which require further confirmation, thus forming a unified set of traffic description fragments corresponding to the target traffic object.

[0027] Furthermore, after obtaining the unified traffic description fragment set, the descriptive relationships of the target traffic object across different data sources are analyzed. On one hand, based on the preceding and following descriptions, partial descriptions, and supplementary descriptions of the same target traffic object in different data sources, a state transfer relationship is established between the different data sources to reflect the state continuity of the target traffic object in cross-source data. On the other hand, based on the differences in information provided by different data sources for the same target traffic object, a semantic completion relationship is established to clarify which data sources can supplement the descriptive content not yet covered by other data sources. On this basis, for the missing, unconfirmed, or strengthened parts in the current descriptions of different target traffic objects, a set of multi-source collaborative data collection instructions corresponding to the target traffic object can be generated to guide different data sources to conduct collaborative data collection around the same target traffic object in the future.

[0028] This step first performs cross-source alignment and conflict resolution on the multi-source traffic data set based on the dynamic confidence representation results. This process maps, filters, and unifies the scattered descriptions of the same target traffic object from different collection sources, forming a unified set of traffic description fragments. Building upon this, it further establishes state transfer relationships and semantic completion relationships for the target traffic object among different collection sources. This ensures that different collection sources not only form consistent descriptions around the same target traffic object but also clarify the supplementary directions and connections between each source. Subsequently, by generating a set of multi-source collaborative collection instructions corresponding to the target traffic object, subsequent collection processes can be carried out in a targeted manner around the unified description results. Therefore, this step provides a stronger cross-source collaborative foundation and higher consistency description capabilities for multi-source road traffic information collection.

[0029] S5: Based on the multi-source collaborative acquisition instruction set, control the corresponding acquisition source to perform adaptive acquisition processing on the target traffic object, and perform back-injection update on the incremental traffic data returned after adaptive acquisition processing, so as to update the unified traffic description fragment set.

[0030] It should be noted that adaptive data acquisition and processing refers to the process by which the data acquisition source dynamically adjusts the acquisition method, content, or timing of acquisition based on a set of multi-source collaborative acquisition instructions for the target traffic object. Incremental traffic data refers to traffic-related data newly acquired by the data acquisition source relative to the existing unified traffic description fragment set after adaptive data acquisition and processing. Injection and update refers to the process of supplementing, correcting, or updating the original description content by writing incremental traffic data into the existing unified traffic description fragment set.

[0031] Specifically, firstly, based on a multi-source collaborative acquisition instruction set, the corresponding acquisition sources are controlled to conduct adaptive acquisition processing around the target traffic object. Since the multi-source collaborative acquisition instruction set already reflects the collaborative needs of different acquisition sources around the target traffic object, each acquisition source can adjust its subsequent acquisition process according to the corresponding instructions. This adjustment can be reflected in the continuous tracking, supplementary observation, or targeted acquisition of the target traffic object, so that the acquisition sources no longer output data independently in a fixed manner, but instead conduct object-oriented acquisition around the description gaps, unconfirmed content, or supplementary content of the current target traffic object, thereby obtaining incremental traffic data further relevant to the target traffic object.

[0032] Furthermore, after obtaining the incremental traffic data, it is further processed by matching this data with the existing unified traffic description fragment set. Specifically, new states, new associations, or newly added descriptive content reflecting the target traffic object in the incremental traffic data are matched with corresponding existing description fragments in the unified traffic description fragment set. Information that can supplement existing descriptions, correct original descriptions, or enrich the semantics of the object is then injected back into the unified traffic description fragment set. Through this injection and update process, the unified traffic description fragment set is no longer a static result formed in one go, but can be continuously updated and expanded with subsequent data collection.

[0033] This step controls the corresponding acquisition sources to adaptively acquire and process the target traffic object based on a multi-source collaborative acquisition command set. This allows subsequent acquisition processes to be targeted around the existing descriptions of the target traffic object, thereby obtaining incremental traffic data that is further relevant to the target traffic object. Furthermore, by injecting this incremental traffic data back into the unified traffic description fragment set, the description of the target traffic object in the unified traffic description fragment set can be continuously supplemented, corrected, and improved. Therefore, this step enables the multi-source road traffic information acquisition process to form a complete processing mechanism of "collaborative acquisition—incremental supplementation—unified update," thereby improving the continuity, completeness, and consistency of the description of the target traffic object.

[0034] This embodiment acquires raw traffic data from at least two heterogeneous data sources in a road traffic scenario. Based on the source identification information, acquisition time sequence information, and spatial location description information corresponding to each data source, the raw traffic data is correlated and arranged to obtain a multi-source traffic data set. Based on the multi-source traffic data set, data quality characterization information, scene state characterization information, and object association clue information corresponding to each raw traffic data are extracted. Based on the data quality characterization information, scene state characterization information, and object association clue information, a dynamic confidence characterization result corresponding to each data source is constructed. Based on the dynamic confidence characterization result, the multi-source traffic data set is then... Cross-source alignment and conflict resolution processes are used to obtain a unified traffic description fragment set corresponding to the target traffic object, where the target traffic object includes at least one of road infrastructure, traffic participants, and traffic events. Based on the unified traffic description fragment set, state transfer relationships and semantic completion relationships of the target traffic object are established among different acquisition sources, generating a multi-source collaborative acquisition instruction set corresponding to the target traffic object. Based on the multi-source collaborative acquisition instruction set, the corresponding acquisition sources are controlled to perform adaptive acquisition processing on the target traffic object, and the incremental traffic data returned after adaptive acquisition processing is back-injected to update the unified traffic description fragment set. This embodiment performs correlation and arrangement processing on raw traffic data from at least two heterogeneous collection sources, establishing a foundation for correlation between data from different collection sources in terms of source, temporal sequence, and spatial dimensions. Furthermore, by extracting data quality characterization information, scene state characterization information, and object association clues, it constructs dynamic confidence characterization results corresponding to each collection source, providing a basis for subsequent processing of multi-source data. Based on this, cross-source alignment and conflict discrimination processing are performed on the multi-source traffic data set to form a unified traffic description fragment set corresponding to the target traffic object. Furthermore, it establishes state transmission relationships and semantic completion relationships between different collection sources for the target traffic object, generating a multi-source collaborative collection instruction set. This enables different collection sources to collaboratively collect data around the same target traffic object. Combined with the injection and updating of incremental traffic data, the unified traffic description fragment set can be continuously supplemented and corrected, thereby improving cross-source collaboration and description consistency in the process of collecting multi-source road traffic information.

[0035] Based on the first embodiment described above, a second embodiment of the road traffic multi-source information collection method of this application is proposed. Please refer to... Figure 2 , Figure 2 This is a schematic diagram of a sub-process in the second embodiment of the road traffic multi-source information collection method of this application.

[0036] like Figure 2 As shown, in this embodiment, step S1 includes: S11: Obtain the raw traffic data output by each heterogeneous acquisition source in the road traffic scene, and determine the acquisition source category and data arrival status corresponding to each raw traffic data. S12: Based on the collection source category and data arrival status corresponding to each original traffic data, determine the source identification information, collection time sequence information and spatial location description information corresponding to each original traffic data, and establish the association relationship between each original traffic data and the source identification information, the collection time sequence information and the spatial location description information; S13: Based on the aforementioned correlation, cross-source aggregation, order arrangement, and location mapping are performed on each original traffic data to obtain a multi-source traffic data set.

[0037] It should be noted that "collection source category" refers to the classification identifier of the type to which each heterogeneous collection source belongs. "Data arrival status" refers to the arrival status of the raw traffic data during access processing. "Cross-source aggregation" refers to the process of aggregating and organizing raw traffic data from different collection sources based on established relationships. "Order arrangement" refers to the process of sequentially organizing and arranging raw traffic data based on collection time sequence information or data arrival status. "Location mapping processing" refers to the process of organizing raw traffic data according to its corresponding spatial region, location unit, or location relationship based on spatial location description information.

[0038] Specifically, firstly, in a road traffic scenario, raw traffic data from various heterogeneous data acquisition sources is received, and the source category and arrival status of each piece of raw traffic data are identified. Since different data acquisition sources typically differ in data access methods, output frequency, and data structure, after the raw traffic data is accessed, its source type is distinguished, and its current arrival status is confirmed to clarify the source attributes and access sequence of different raw traffic data. Based on this, corresponding source identification information, acquisition time sequence information, and spatial location description information are further determined for each piece of raw traffic data, enabling each piece of raw traffic data to be clearly marked in the source, time, and spatial dimensions, and establishing the association between each piece of raw traffic data and the aforementioned information.

[0039] Furthermore, after establishing the correlation, the original traffic data is then organized and processed based on this correlation. Specifically, original traffic data from different sources but with correlation can first be aggregated across sources, bringing related data together under a unified data organization framework. Then, based on the collection time sequence information and data arrival status, the aggregated original traffic data is arranged in order to form a temporally correlated arrangement. Simultaneously, based on spatial location description information, each original traffic data is mapped to a corresponding road area, traffic node, or scene location. After the above processing, a multi-source traffic data set with source correlation, temporal correlation, and spatial correlation can be obtained.

[0040] This step first identifies the data source category and arrival status of each piece of raw traffic data, and further determines the source identification information, collection time sequence information, and spatial location description information. It then establishes the correlation between each piece of raw traffic data and the aforementioned information, thus providing a unified organizational foundation for the dispersed raw traffic data at the source, time, and spatial levels. Based on this, cross-source aggregation, sequential arrangement, and location mapping are then used to form a multi-source traffic data set, enabling the orderly aggregation and corresponding organization of raw traffic data output from different data sources. Therefore, this step provides a structured data foundation for subsequent analysis and processing of multi-source traffic data.

[0041] In this embodiment, step S2 includes: S21: Based on the multi-source traffic data set, each original traffic data is parsed and processed to determine the data integrity information, data consistency information and data stability information corresponding to each original traffic data, so as to generate data quality characterization information corresponding to each original traffic data. S22: Based on the multi-source traffic data set, scene element identification and object relationship identification are performed on each original traffic data to determine the traffic flow state information, road environment state information and object correspondence information corresponding to each original traffic data, so as to generate scene state representation information and object association clue information corresponding to each original traffic data. S23: Based on the data quality characterization information, scene state characterization information and object association clue information corresponding to each original traffic data, the data credibility and association support of each collection source are jointly characterized, and a dynamic confidence characterization result corresponding to each collection source is constructed.

[0042] Step S23 includes: S231: Based on the data quality characterization information corresponding to each original traffic data, determine the data reliability evaluation information corresponding to each collection source in the current collection cycle, wherein the data reliability evaluation information is used to characterize the usability and fluctuation status of the output data of each collection source; S232: Based on the scene state representation information and object association clue information corresponding to each original traffic data, determine the association support evaluation information of each collection source for the target traffic object and the target traffic scene, wherein the association support evaluation information is used to characterize the degree of support of each collection source for the traffic object identification result and the scene state determination result. S233: Based on the data credibility evaluation information and the correlation support evaluation information corresponding to each collection source, the data credibility and correlation support of each collection source are jointly characterized to construct a dynamic confidence representation result corresponding to each collection source.

[0043] It should be noted that data integrity information refers to information used to characterize whether the content of the original traffic data is complete, whether any fields are missing, and whether the description is complete. Data consistency information refers to information used to characterize whether the content expression, state indication, or result description is consistent within the same original traffic data or between different original traffic data. Data stability information refers to information used to characterize whether the fluctuation of the original traffic data is stable during continuous collection or multiple outputs. Scene element identification refers to the process of identifying state elements, environmental elements, or operational elements related to the road traffic scene from the original traffic data. Object relationship identification refers to the process of identifying the correspondence, association, or interaction relationships between different traffic objects from the original traffic data. Traffic flow state information refers to information used to characterize the operational state of traffic flow in the road traffic scene. Road environment state information refers to information used to characterize the environmental conditions, road state, or external scene state in the road traffic scene. Object correspondence information refers to information used to characterize whether the objects described in different original traffic data point to the same target object or have an association relationship. Data reliability evaluation information refers to evaluation information determined based on data quality characterization information, used to characterize the usability and fluctuation state of the output data from each collection source. Association support evaluation information refers to the evaluation information determined based on scene state representation information and object association clue information, used to characterize the degree of support of each data collection source for the traffic object identification results and scene state determination results. The current data collection cycle refers to the time range corresponding to the current round of data collection, reception, and processing for each data collection source.

[0044] Specifically, firstly, based on a multi-source traffic dataset, the original traffic data is analyzed in detail. This analysis focuses on the data state of the original traffic data itself, determining its data integrity, consistency, and stability information to generate data quality characterization information for each piece of original traffic data. Data integrity information reflects whether there are missing or incomplete contents in the original traffic data; data consistency information reflects whether there are contradictions or inconsistencies between the original traffic data and other related data; and data stability information reflects whether there are abnormal fluctuations in the original traffic data during continuous output. Simultaneously, based on the multi-source traffic dataset, scene element identification and object relationship identification can be performed on each piece of original traffic data to further determine the traffic flow state information, road environment state information, and object correspondence information corresponding to each piece of original traffic data, thereby generating scene state characterization information and object association clue information for each piece of original traffic data. Through the above processing, not only can the data quality characterization results of each piece of original traffic data be obtained, but also the scene state and object association information they reflect can be obtained.

[0045] Furthermore, based on this, the above information is further summarized and jointly analyzed at the data collection source level. Specifically, firstly, based on the data quality characterization information corresponding to each original traffic data, the data credibility evaluation information corresponding to each data collection source in the current collection cycle can be determined, thereby reflecting the usability and fluctuation status of the output data of each data collection source in the current stage. Secondly, based on the scene state characterization information and object association clue information corresponding to each original traffic data, the association support evaluation information of each data collection source for the target traffic object and the target traffic scene can be determined, so as to reflect the degree of support of each data collection source for the object identification results and scene judgment results. Subsequently, the data credibility evaluation information and association support evaluation information corresponding to each data collection source are jointly characterized, thereby constructing a dynamic confidence characterization result corresponding to each data collection source. This dynamic confidence characterization result can comprehensively reflect the data credibility of different data collection sources in the current collection cycle and their support capability for the description of traffic objects and traffic scenes.

[0046] This step first analyzes and processes the raw traffic data to generate data quality characterization information. Then, it further generates scene state characterization information and object association clue information through scene element identification and object relationship identification, thus characterizing the raw traffic data from three levels: the data's own state, the state reflected in the scene, and the state of object association. Based on this, it further determines the data credibility evaluation information and association support evaluation information corresponding to each collection source, and jointly represents these two to construct a dynamic confidence characterization result corresponding to each collection source. Therefore, this step enables different collection sources to be differentiated not only in terms of data availability but also in terms of their support capabilities for target traffic objects and target traffic scenarios, thus providing a more targeted basis for subsequent cross-source alignment, conflict detection, and collaborative collection of multi-source traffic data.

[0047] This embodiment performs correlation and arrangement processing on raw traffic data from at least two heterogeneous collection sources, establishing a foundation for correlation between data from different collection sources in terms of source, temporal sequence, and spatial dimensions. Furthermore, by extracting data quality characterization information, scene state characterization information, and object association clues, it constructs dynamic confidence characterization results corresponding to each collection source, providing a basis for subsequent processing of multi-source data. Based on this, cross-source alignment and conflict discrimination processing are performed on the multi-source traffic data set to form a unified traffic description fragment set corresponding to the target traffic object. Furthermore, it establishes state transmission relationships and semantic completion relationships between different collection sources for the target traffic object, generating a multi-source collaborative collection instruction set. This enables different collection sources to collaboratively collect data around the same target traffic object. Combined with the injection and updating of incremental traffic data, the unified traffic description fragment set can be continuously supplemented and corrected, thereby improving cross-source collaboration and description consistency in the process of collecting multi-source road traffic information.

[0048] Based on the second embodiment described above, a third embodiment of the road traffic multi-source information collection method of this application is proposed. Please refer to... Figure 3 , Figure 3 This is a schematic diagram of a sub-process in the third embodiment of the road traffic multi-source information collection method of this application.

[0049] In this embodiment, step S3 includes: S31: Based on the dynamic confidence characterization results, determine the confidence participation relationship and alignment constraint relationship corresponding to each original traffic data in the multi-source traffic data set, and filter candidate traffic data to participate in cross-source fusion according to the confidence participation relationship and the alignment constraint relationship; S32: Based on the candidate traffic data participating in cross-source fusion, the object identification information, status description information and location association information pointing to the same target traffic object in different collection sources are matched and aligned across sources to obtain the alignment description results corresponding to each target traffic object; S33: Based on the alignment description results corresponding to each target traffic object, and combined with the description differences between different collection sources, conflict-causing description content is processed for conflict discrimination, and the valid description content after discrimination is merged and organized to obtain a unified traffic description fragment set corresponding to the target traffic object.

[0050] It should be noted that: Confidence participation relationship refers to the relationship information determined based on the dynamic confidence representation results, which specifies the degree or mode of participation of each original traffic data in the subsequent cross-source fusion process. Alignment constraint relationship refers to the relationship information used to limit the corresponding conditions or matching rules that different data sources should meet during the cross-source alignment process. Candidate traffic data refers to traffic data identified as participating in cross-source fusion processing after being filtered based on confidence participation relationship and alignment constraint relationship. Object identification information refers to information used to characterize the identity of the target traffic object or the object it points to. Status description information refers to information used to characterize the current status, changes, or operational characteristics of the target traffic object. Location association information refers to information used to characterize the spatial location of the target traffic object and its correspondence with other location descriptions. Alignment description result refers to the result formed after corresponding matching and cross-source alignment processing of relevant descriptions pointing to the same target traffic object from different data sources.

[0051] Specifically, firstly, based on the dynamic confidence representation results, the original traffic data in the multi-source traffic dataset are screened before fusion. Specifically, based on the dynamic confidence status of the collection source corresponding to each original traffic data, it can be determined which original traffic data are suitable for subsequent cross-source fusion, and the confidence participation relationship of each original traffic data in the fusion is further determined. Simultaneously, corresponding alignment constraints can be established by combining the correspondence requirements between different original traffic data in terms of time, space, object orientation, and descriptive semantics. Based on this, according to the confidence participation relationship and alignment constraints, the original traffic data in the multi-source traffic dataset is screened to obtain candidate traffic data for cross-source fusion. Through this process, data that is relevant to the subsequent fusion target and meets the alignment conditions can be retained, establishing an input foundation for the unified description of the same target traffic object in the future.

[0052] Furthermore, after obtaining candidate traffic data, cross-source alignment processing is performed around the same target traffic object. Specifically, object identification information, status description information, and location association information pointing to the same target traffic object from different data sources are matched to identify which descriptions belong to the same object, and these descriptions are then aligned across sources to form aligned descriptions corresponding to each target traffic object. Subsequently, considering the description differences between different data sources, conflict resolution is performed on inconsistent, contradictory, or difficult-to-coexist descriptions to identify valid descriptions that can be retained and used for unified organization. Finally, the identified valid descriptions are merged and organized to form a unified set of traffic description fragments corresponding to the target traffic object, thereby integrating the originally scattered and differentiated object descriptions from different data sources into a unified description result.

[0053] This step first determines the confidence participation relationships and alignment constraints corresponding to each original traffic data based on the dynamic confidence representation results, and then filters candidate traffic data to participate in cross-source fusion, thus ensuring that subsequent fusion processing is based on data with participation evidence and meeting the corresponding conditions. Subsequently, by performing corresponding matching and cross-source alignment on object identification information, status description information, and location association information pointing to the same target traffic object from different collection sources, related descriptions scattered across different collection sources can be uniformly associated. Furthermore, by combining the description differences between different collection sources to determine conflicts and merging and organizing valid description content, a unified set of traffic description fragments corresponding to the target traffic object can be formed. Therefore, this step enhances the ability to uniformly organize multi-source traffic data around the same target traffic object and provides a consistent descriptive foundation for subsequent establishment of cross-source collaborative relationships and collaborative collection.

[0054] Based on the second embodiment described above, in this embodiment, step S4 includes: S41: Based on the unified traffic description fragment set, extract the object state description information, state change clue information and cross-source semantic association information corresponding to each target traffic object, and establish the state correspondence relationship between each target traffic object and different collection sources. S42: Based on the state correspondence relationship of each target traffic object and the cross-source semantic association information, determine the state transmission relationship and semantic completion relationship of each target traffic object between different collection sources, and generate collaborative collection requirement information corresponding to each target traffic object; S43: Based on the collaborative collection requirement information corresponding to each target traffic object, determine the collection collaboration method, collection content and collection timing corresponding to different collection sources, and generate a multi-source collaborative collection instruction set corresponding to the target traffic object.

[0055] It should be noted that object state description information refers to information used to characterize the current state, state attributes, or state performance of a target traffic object. State change clue information refers to information used to reflect the state evolution trend, signs of change, or clues to the connection between previous and subsequent states of a target traffic object. Cross-source semantic association information refers to information used to characterize the semantic correspondence, supplementation, or association between descriptions of the same target traffic object in different data collection sources. State correspondence refers to the correspondence established between various state descriptions of the target traffic object in different data collection sources. Collaborative data collection requirement information refers to the requirement information formed based on the state transmission relationship and semantic completion relationship of the target traffic object between different data collection sources, used to guide subsequent collaborative data collection arrangements. Data collection collaboration method refers to the cooperation method or collaborative organization method adopted by different data collection sources when conducting collaborative data collection around the target traffic object.

[0056] Specifically, firstly, based on a unified traffic description fragment set, each target traffic object is further analyzed. Specifically, object state description information, state change clue information, and cross-source semantic association information corresponding to each target traffic object can be extracted from the unified traffic description fragment set. Among these, the object state description information reflects the currently formed state description content of the target traffic object; the state change clue information reflects the change trend and state connection of the target traffic object in the description fragment; and the cross-source semantic association information reflects the semantic association of the same target traffic object by different data collection sources. Based on this, and combining the above information, a state correspondence relationship between each target traffic object and different data collection sources is established, thereby clarifying which state contents correspond to the state descriptions of the same target traffic object in different data collection sources and how they are connected.

[0057] Furthermore, after establishing the state correspondence, the state transmission relationship and semantic completion relationship of each target traffic object are determined between different data collection sources based on the state correspondence of each target traffic object and cross-source semantic association information. In other words, it identifies which data collection sources can reflect the state continuation process of the target traffic object, and which data collection sources can semantically supplement content not fully described by other data collection sources. Based on this, collaborative data collection requirement information corresponding to each target traffic object is generated. Subsequently, based on the collaborative data collection requirement information, the data collection collaboration method, data collection content, and data collection sequence corresponding to different data collection sources are determined, enabling different data collection sources to coordinate and cooperate around the target traffic object in subsequent data collection, and ultimately generating a multi-source collaborative data collection instruction set corresponding to the target traffic object.

[0058] This step extracts object state description information, state change clues, and cross-source semantic association information for each target traffic object based on a unified traffic description fragment set. It also establishes state correspondences between different data collection sources, thus providing a consistent state description foundation for different data collection sources around the same target traffic object. Furthermore, by determining state transmission relationships and semantic completion relationships and generating collaborative data collection requirement information, it clarifies the connection and supplementary directions for different data collection sources in subsequent data collection. Based on this, it determines the collaborative data collection methods, content, and timing corresponding to different data collection sources and generates a multi-source collaborative data collection instruction set, enabling different data collection sources to conduct organized collaborative data collection around the target traffic object. Therefore, this step improves cross-source collaboration and descriptive continuity in the multi-source traffic information collection process.

[0059] In this embodiment, step S5 includes: S51: Based on the multi-source collaborative acquisition instruction set, determine the target acquisition source, target acquisition content, and target acquisition timing corresponding to each target traffic object, and control the corresponding acquisition source to enter the adaptive acquisition state corresponding to the target traffic object; S52: Based on the adaptive acquisition state corresponding to each acquisition source, obtain incremental traffic data corresponding to the target traffic object, and based on the object correspondence and description association between the incremental traffic data and the unified traffic description fragment set, determine the incremental update content corresponding to the target traffic object; S53: Based on the incremental update content corresponding to the target traffic object, perform back-injection update on the unified traffic description fragment set to obtain the updated unified traffic description fragment set.

[0060] Step S51 includes: S511: Based on the multi-source collaborative acquisition instruction set, analyze the acquisition collaboration requirements corresponding to each target traffic object, and determine the candidate acquisition source set, candidate acquisition content set, and candidate acquisition time sequence set corresponding to each target traffic object; S512: Based on the candidate collection source set, the candidate collection content set, and the candidate collection time sequence set corresponding to each target traffic object, perform corresponding matching and constraint filtering processing to determine the target collection source, target collection content, and target collection time sequence corresponding to each target traffic object; S513: Based on the target acquisition source, the target acquisition content, and the target acquisition timing corresponding to each target traffic object, send acquisition control information to the corresponding acquisition source and control the corresponding acquisition source to enter the adaptive acquisition state corresponding to the target traffic object.

[0061] It should be noted that the collection coordination requirements refer to the collection cooperation conditions, collection direction requirements, or collection organization requirements specified for each target traffic object in the multi-source collaborative collection instruction set. The candidate collection source set refers to the set of collection sources initially determined based on the collection coordination requirements, which can be used to perform subsequent collection on the target traffic object. The candidate collection timing set refers to the set of time arrangements related to the subsequent collection of the target traffic object, initially determined based on the collection coordination requirements. Collection control information refers to control information used to control the entry of the corresponding collection source into the target collection process. Object correspondence refers to the correspondence established between incremental traffic data and the unified traffic description fragment set at the target traffic object level. Description association refers to the association established between the description content in the incremental traffic data and the existing description content in the unified traffic description fragment set. Incremental update content refers to the new or adjusted content used to update the unified traffic description fragment set, determined based on the object correspondence and description association between the incremental traffic data and the unified traffic description fragment set. The updated unified traffic description fragment set refers to the set of description results formed after the back-injection update based on the original unified traffic description fragment set.

[0062] Specifically, firstly, based on the multi-source collaborative acquisition instruction set, the acquisition coordination requirements corresponding to each target traffic object are analyzed. Through analysis, the candidate acquisition source set, candidate acquisition content set, and candidate acquisition time sequence set corresponding to each target traffic object can be determined. That is, it is first clarified which acquisition sources have the potential to participate in subsequent acquisition, which content belongs to the scope of supplementary or key acquisition, and which time arrangements are suitable for acquisition around the target traffic object. Subsequently, based on the candidate acquisition source set, candidate acquisition content set, and candidate acquisition time sequence set corresponding to each target traffic object, corresponding matching and constraint screening processing is performed to determine the target acquisition source, target acquisition content, and target acquisition time sequence corresponding to each target traffic object. After the determination is completed, acquisition control information is further issued to the corresponding acquisition source based on the target acquisition source, target acquisition content, and target acquisition time sequence corresponding to each target traffic object, and the corresponding acquisition source is controlled to enter the adaptive acquisition state corresponding to the target traffic object, so that different acquisition sources can carry out targeted subsequent acquisition around the target traffic object.

[0063] Furthermore, after each data source enters the adaptive data acquisition state, incremental traffic data for the corresponding target traffic object is acquired. Since the data acquisition process is now organized around the target traffic object, the acquired incremental traffic data can more specifically reflect the new state, new associated information, or supplementary descriptive content of that target traffic object. Next, based on the object correspondence and descriptive association between the incremental traffic data and the unified traffic description fragment set, the incremental update content corresponding to the target traffic object is determined. That is, it is clarified which newly acquired data should be added to the original description, which content should be used to correct existing descriptions, and which content can be used to improve object semantics or state chains. Finally, based on the incremental update content corresponding to the target traffic object, the unified traffic description fragment set is back-injected and updated to obtain the updated unified traffic description fragment set, thus enabling the unified description results to continuously evolve and improve with subsequent data acquisition.

[0064] This step first analyzes the data collection coordination requirements corresponding to each target traffic object to determine the candidate data collection source set, candidate data collection content set, and candidate data collection time sequence set. It then further filters and determines the target data collection source, target data collection content, and target data collection time sequence, enabling the subsequent data collection process to be targeted towards the target traffic object. Next, by controlling the corresponding data collection source to enter an adaptive data collection state, incremental traffic data corresponding to the target traffic object can be acquired. Combining the object correspondence and description association between the incremental traffic data and the unified traffic description fragment set, the incremental update content is determined. Finally, by back-injecting and updating the unified traffic description fragment set, the updated unified traffic description fragment set is obtained. Therefore, this step enables the multi-source traffic information collection process to form a continuous supplementation and dynamic update mechanism oriented towards the target traffic object, thereby improving the continuity, completeness, and consistency of the unified traffic description results.

[0065] This embodiment performs correlation and arrangement processing on raw traffic data from at least two heterogeneous collection sources, establishing a foundation for correlation between data from different collection sources in terms of source, temporal sequence, and spatial dimensions. Furthermore, by extracting data quality characterization information, scene state characterization information, and object association clues, it constructs dynamic confidence characterization results corresponding to each collection source, providing a basis for subsequent processing of multi-source data. Based on this, cross-source alignment and conflict discrimination processing are performed on the multi-source traffic data set to form a unified traffic description fragment set corresponding to the target traffic object. Furthermore, it establishes state transmission relationships and semantic completion relationships between different collection sources for the target traffic object, generating a multi-source collaborative collection instruction set. This enables different collection sources to collaboratively collect data around the same target traffic object. Combined with the injection and updating of incremental traffic data, the unified traffic description fragment set can be continuously supplemented and corrected, thereby improving cross-source collaboration and description consistency in the process of collecting multi-source road traffic information.

[0066] In one embodiment, to address the problem that short-term abnormal events in road traffic scenarios are difficult to continuously perceive and that descriptions of abnormal events from different data sources are prone to breakage, an event trigger memory unit and an abnormal backtracking chain generation module can be further set up on the basis of the above-mentioned multi-source information collection method for road traffic.

[0067] Specifically, after associating and arranging the raw traffic data to form a multi-source traffic data set, event trigger detection can be performed on abrupt state segments, abnormal state segments, or segments with sudden increases in correlation strength within the multi-source traffic data set. When a target segment that meets preset trigger conditions is detected, the raw traffic data, acquisition source identification information, acquisition time sequence information, spatial location description information, and corresponding dynamic confidence representation results corresponding to the target segment are written into the event trigger memory unit to form an event memory segment corresponding to the target segment. The event trigger memory unit can be used to retain the contextual description content of the abnormal event in multiple correlation stages before and after triggering, so that subsequent processing is not only oriented towards the acquisition results at the current moment, but also able to call historical correlation information in the formation process of the abnormal event.

[0068] Furthermore, after forming a unified set of traffic description fragments, the associated descriptions of the same target traffic object under different time periods, different collection sources, and different state stages can be backtracked and linked based on the event memory fragments stored in the event trigger memory unit to generate a corresponding anomaly backtracking chain. The anomaly backtracking chain can characterize the state evolution path of the target traffic object before the anomaly occurs, when the anomaly appears, and during the anomaly continuation stage, and record the description access status and semantic supplementation relationships of different collection sources at each stage. Based on the anomaly backtracking chain, the breakpoint locations, information sparse locations, and semantic conflict concentration locations in the anomaly description can be further identified, and the identification results can be used as additional constraints input into the multi-source collaborative collection instruction generation process, so that subsequent collection sources prioritize supplementing the collection of breakpoint segments and weak segments in the anomaly backtracking chain. In some embodiments, a corresponding cross-stage association index relationship can also be constructed based on the anomaly backtracking chain to chain together the unified traffic description fragment sets of the same target traffic object under different collection cycles. In this way, when incremental traffic data is back-injected and updated in the future, not only can the current unified traffic description fragment set be updated, but also the relevant chain segments in the anomaly backtracking chain corresponding to the target traffic object can be corrected simultaneously, so that the description of the anomaly evolution process of the target traffic object is consistent with the current unified description result.

[0069] Through the above implementation methods, the system's ability to continuously retain and retrospectively organize the formation process of abnormal traffic events can be further enhanced on the basis of the original multi-source traffic information collection, cross-source alignment, unified description and collaborative collection, thereby providing more detailed support for the subsequent supplementary collection, status verification and continuous description of abnormal traffic events.

[0070] This application also provides a road traffic multi-source information collection device. Please refer to... Figure 4 , Figure 4 This is a schematic diagram of the module structure of a road traffic multi-source information collection device according to an embodiment of this application. The road traffic multi-source information collection device includes: The data acquisition module 401 is used to acquire raw traffic data from at least two heterogeneous acquisition sources in a road traffic scenario, and to perform association and arrangement processing on the raw traffic data based on the source identification information, acquisition time sequence information and spatial location description information corresponding to each acquisition source, to obtain a multi-source traffic data set. The confidence characterization module 402 is used to extract data quality characterization information, scene state characterization information and object association clue information corresponding to each original traffic data based on the multi-source traffic data set, and to construct a dynamic confidence characterization result corresponding to each collection source based on the data quality characterization information, the scene state characterization information and the object association clue information. The unified description module 403 is used to perform cross-source alignment processing and conflict discrimination processing on the multi-source traffic data set based on the dynamic confidence representation results, so as to obtain a unified traffic description fragment set corresponding to the target traffic object, wherein the target traffic object includes at least one of road infrastructure, traffic participants and traffic events. The multi-source collaboration module 404 is used to establish the state transmission relationship and semantic completion relationship of the target traffic object among different acquisition sources based on the unified traffic description fragment set, and generate a multi-source collaborative acquisition instruction set corresponding to the target traffic object. The injection update module 405 is used to control the corresponding acquisition source to perform adaptive acquisition processing on the target traffic object based on the multi-source collaborative acquisition instruction set, and to inject and update the incremental traffic data returned after adaptive acquisition processing in order to update the unified traffic description fragment set.

[0071] The road traffic multi-source information acquisition device provided in this application adopts the road traffic multi-source information acquisition method in the above embodiments, which can solve the technical problem of how to improve cross-source coordination and description consistency in the road traffic multi-source information acquisition process. Compared with the prior art, the beneficial effects of the road traffic multi-source information acquisition device provided in this application are the same as the beneficial effects of the road traffic multi-source information acquisition method provided in the above embodiments, and other technical features in the road traffic multi-source information acquisition device are the same as the features disclosed in the methods of the above embodiments, and will not be repeated here.

[0072] This application provides a road traffic multi-source information acquisition device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the road traffic multi-source information acquisition method described above.

[0073] The following is for reference. Figure 5 , Figure 5 This is a schematic diagram of the hardware operating environment of the road traffic multi-source information collection method in the embodiments of this application, showing a schematic diagram of the structure of the road traffic multi-source information collection device suitable for implementing the embodiments of this application. Figure 5 The road traffic multi-source information collection device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0074] like Figure 5As shown, the road traffic multi-source information acquisition device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the road traffic multi-source information acquisition device. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the road traffic multi-source information acquisition equipment to communicate wirelessly or wiredly with other devices to exchange data. Although the figure shows a road traffic multi-source information acquisition equipment with various systems, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems can be implemented alternatively.

[0075] In particular, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. When the computer program is executed by the processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0076] The road traffic multi-source information collection device provided in this application, employing the road traffic multi-source information collection method described in the above embodiments, can solve the technical problem of how to improve cross-source coordination and description consistency in the road traffic multi-source information collection process. Compared with the prior art, the beneficial effects of the road traffic multi-source information collection device provided in this application are the same as those of the road traffic multi-source information collection method provided in the above embodiments, and other technical features in this road traffic multi-source information collection device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0077] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0078] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0079] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, which are used to execute the road traffic multi-source information collection method described in the above embodiments.

[0080] The aforementioned computer-readable storage medium carries one or more programs. When these programs are executed by the road traffic multi-source information acquisition device, the road traffic multi-source information acquisition device: acquires raw traffic data from at least two types of heterogeneous acquisition sources in a road traffic scene; and performs association and arrangement processing on the raw traffic data based on the source identification information, acquisition time sequence information, and spatial location description information corresponding to each acquisition source to obtain a multi-source traffic data set; based on the multi-source traffic data set, extracts data quality characterization information, scene state characterization information, and object association clue information corresponding to each raw traffic data; and based on the data quality characterization information, scene state characterization information, and object association clue information, constructs a dynamic association system corresponding to each acquisition source. The dynamic confidence representation results are used to perform cross-source alignment and conflict discrimination processing on the multi-source traffic data set, resulting in a unified traffic description fragment set corresponding to the target traffic object. The target traffic object includes at least one of road infrastructure, traffic participants, and traffic events. Based on the unified traffic description fragment set, state transfer relationships and semantic completion relationships of the target traffic object are established between different acquisition sources, generating a multi-source collaborative acquisition instruction set corresponding to the target traffic object. Based on the multi-source collaborative acquisition instruction set, the corresponding acquisition sources are controlled to perform adaptive acquisition processing on the target traffic object, and the incremental traffic data returned after adaptive acquisition processing is back-injected to update the unified traffic description fragment set. Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof. These programming languages ​​include object-oriented programming languages—such as Java, Smalltalk, and C++—and conventional procedural programming languages—such as the "C" language or similar programming languages. The program code can execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0081] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0082] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0083] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described road traffic multi-source information acquisition method. This addresses the technical problem of improving cross-source coordination and description consistency during the road traffic multi-source information acquisition process. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the road traffic multi-source information acquisition method provided in the above embodiments, and will not be elaborated upon here.

[0084] This application provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the road traffic multi-source information collection method described above.

[0085] The computer program product provided in this application can solve the technical problem of how to improve cross-source coordination and description consistency in the process of collecting multi-source information on road traffic. Compared with the prior art, the beneficial effects of the computer program product provided in the embodiments of this application are the same as the beneficial effects of the road traffic multi-source information collection method provided in the above embodiments, and will not be repeated here.

[0086] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the scope of protection of the present invention. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the essence and scope of the technical solutions of the present invention.

Claims

1. A method for collecting multi-source road traffic information, characterized in that, The method includes: The system acquires raw traffic data from at least two heterogeneous data sources in a road traffic scenario, and performs association and arrangement processing on the raw traffic data based on the source identification information, acquisition time sequence information and spatial location description information corresponding to each data source to obtain a multi-source traffic data set. Based on the multi-source traffic data set, data quality characterization information, scene state characterization information, and object association clue information corresponding to each original traffic data are extracted. Based on the data quality characterization information, scene state characterization information, and object association clue information, dynamic confidence characterization results corresponding to each data collection source are constructed. Based on the dynamic confidence representation results, cross-source alignment and conflict discrimination processing are performed on the multi-source traffic data set to obtain a unified traffic description fragment set corresponding to the target traffic object, wherein the target traffic object includes at least one of road infrastructure, traffic participants and traffic events. Based on the unified traffic description fragment set, the state transmission relationship and semantic completion relationship of the target traffic object among different collection sources are established, and a multi-source collaborative collection instruction set corresponding to the target traffic object is generated. Based on the multi-source collaborative acquisition instruction set, the corresponding acquisition source is controlled to perform adaptive acquisition processing on the target traffic object, and the incremental traffic data returned after adaptive acquisition processing is injected back to update the unified traffic description fragment set.

2. The method as described in claim 1, characterized in that, The step of acquiring raw traffic data from at least two heterogeneous data sources in a road traffic scenario, and performing association and arrangement processing on the raw traffic data based on the source identification information, acquisition time sequence information, and spatial location description information corresponding to each data source to obtain a multi-source traffic data set, includes: Acquire raw traffic data output from various heterogeneous data acquisition sources in a road traffic scenario, and determine the data acquisition source category and data arrival status corresponding to each raw traffic data. Based on the source category and arrival status of each original traffic data, the source identification information, collection time sequence information and spatial location description information corresponding to each original traffic data are determined, and the association between each original traffic data and the source identification information, the collection time sequence information and the spatial location description information is established. Based on the aforementioned relationships, the original traffic data are processed through cross-source aggregation, sequential arrangement, and location mapping to obtain a multi-source traffic data set.

3. The method as described in claim 1, characterized in that, The step of extracting data quality characterization information, scene state characterization information, and object association clue information corresponding to each original traffic data based on the multi-source traffic data set, and constructing dynamic confidence characterization results corresponding to each data collection source based on the data quality characterization information, the scene state characterization information, and the object association clue information, includes: Based on the multi-source traffic data set, each original traffic data is parsed and processed to determine the data integrity information, data consistency information, and data stability information corresponding to each original traffic data, so as to generate data quality characterization information corresponding to each original traffic data. Based on the multi-source traffic data set, scene element identification and object relationship identification are performed on each original traffic data to determine the traffic flow state information, road environment state information and object correspondence information corresponding to each original traffic data, so as to generate scene state representation information and object association clue information corresponding to each original traffic data. Based on the data quality characterization information, scene state characterization information, and object association clue information corresponding to each original traffic data, the data credibility and association support of each collection source are jointly characterized, and a dynamic confidence characterization result corresponding to each collection source is constructed.

4. The method as described in claim 3, characterized in that, The step of jointly characterizing the data credibility and correlation support of each collection source based on the data quality characterization information, scene state characterization information, and object association clue information corresponding to each original traffic data, and constructing a dynamic confidence characterization result corresponding to each collection source, includes: Based on the data quality characterization information corresponding to each original traffic data, the data reliability evaluation information corresponding to each collection source in the current collection cycle is determined, wherein the data reliability evaluation information is used to characterize the usability and fluctuation status of the output data of each collection source. Based on the scene state representation information and object association clue information corresponding to each original traffic data, the association support evaluation information of each collection source for the target traffic object and the target traffic scene is determined. The association support evaluation information is used to characterize the degree of support of each collection source for the traffic object identification result and the scene state determination result. Based on the data credibility evaluation information and the correlation support evaluation information corresponding to each collection source, the data credibility and correlation support of each collection source are jointly characterized to construct a dynamic confidence representation result corresponding to each collection source.

5. The method as described in claim 1, characterized in that, The step of performing cross-source alignment and conflict discrimination processing on the multi-source traffic data set based on the dynamic confidence representation results to obtain a unified traffic description fragment set corresponding to the target traffic object includes: Based on the dynamic confidence representation results, the confidence participation relationship and alignment constraint relationship corresponding to each original traffic data in the multi-source traffic data set are determined, and candidate traffic data to participate in cross-source fusion are screened according to the confidence participation relationship and the alignment constraint relationship. Based on the candidate traffic data participating in cross-source fusion, the object identification information, status description information and location association information pointing to the same target traffic object in different collection sources are matched and aligned across sources to obtain the alignment description results corresponding to each target traffic object. Based on the alignment description results corresponding to each target traffic object, and combined with the description differences between different collection sources, conflicting description content is processed for conflict discrimination, and the valid description content after discrimination is merged and organized to obtain a unified traffic description fragment set corresponding to the target traffic object.

6. The method as described in claim 1, characterized in that, The step of establishing the state transfer relationship and semantic completion relationship of the target traffic object among different data collection sources based on the unified traffic description fragment set, and generating a multi-source collaborative data collection instruction set corresponding to the target traffic object, includes: Based on the unified traffic description fragment set, extract the object state description information, state change clue information and cross-source semantic association information corresponding to each target traffic object, and establish the state correspondence relationship between each target traffic object and different collection sources. Based on the state correspondence of each target traffic object and the cross-source semantic association information, the state transmission relationship and semantic completion relationship of each target traffic object between different collection sources are determined, and collaborative collection requirement information corresponding to each target traffic object is generated. Based on the collaborative collection requirement information corresponding to each target traffic object, the collaborative collection method, collection content and collection timing corresponding to different collection sources are determined, and a set of multi-source collaborative collection instructions corresponding to the target traffic object is generated.

7. The method as described in claim 1, characterized in that, The step of controlling the corresponding acquisition source to perform adaptive acquisition processing on the target traffic object based on the multi-source collaborative acquisition instruction set, and injecting back the incremental traffic data returned after adaptive acquisition processing to update the unified traffic description fragment set, includes: Based on the multi-source collaborative acquisition instruction set, the target acquisition source, target acquisition content, and target acquisition timing corresponding to each target traffic object are determined, and the corresponding acquisition source is controlled to enter the adaptive acquisition state corresponding to the target traffic object. Based on the adaptive acquisition state corresponding to each acquisition source, incremental traffic data corresponding to the target traffic object is obtained, and based on the object correspondence and description association between the incremental traffic data and the unified traffic description fragment set, the incremental update content corresponding to the target traffic object is determined. Based on the incremental update content corresponding to the target traffic object, the unified traffic description fragment set is back-injected and updated to obtain the updated unified traffic description fragment set.

8. The method as described in claim 7, characterized in that, The step of determining the target acquisition source, target acquisition content, and target acquisition timing corresponding to each target traffic object based on the multi-source collaborative acquisition instruction set, and controlling the corresponding acquisition source to enter the adaptive acquisition state corresponding to the target traffic object, includes: Based on the multi-source collaborative acquisition instruction set, the acquisition collaboration requirements corresponding to each target traffic object are analyzed, and the candidate acquisition source set, candidate acquisition content set, and candidate acquisition time sequence set corresponding to each target traffic object are determined. Based on the candidate collection source set, the candidate collection content set, and the candidate collection time sequence set corresponding to each target traffic object, corresponding matching and constraint filtering processes are performed to determine the target collection source, target collection content, and target collection time sequence corresponding to each target traffic object. Based on the target acquisition source, the target acquisition content, and the target acquisition timing corresponding to each target traffic object, acquisition control information is sent to the corresponding acquisition source, and the corresponding acquisition source is controlled to enter the adaptive acquisition state corresponding to the target traffic object.

9. A road traffic multi-source information collection device, characterized in that, The device includes: The data acquisition module is used to acquire raw traffic data from at least two heterogeneous collection sources in a road traffic scenario, and to perform association and arrangement processing on the raw traffic data based on the source identification information, collection time sequence information and spatial location description information corresponding to each collection source to obtain a multi-source traffic data set. The confidence characterization module is used to extract data quality characterization information, scene state characterization information, and object association clue information corresponding to each original traffic data based on the multi-source traffic data set, and to construct dynamic confidence characterization results corresponding to each collection source based on the data quality characterization information, the scene state characterization information, and the object association clue information. The unified description module is used to perform cross-source alignment and conflict discrimination processing on the multi-source traffic data set based on the dynamic confidence representation results, so as to obtain a unified traffic description fragment set corresponding to the target traffic object, wherein the target traffic object includes at least one of road infrastructure, traffic participants and traffic events. The multi-source collaboration module is used to establish the state transmission relationship and semantic completion relationship of the target traffic object among different acquisition sources based on the unified traffic description fragment set, and generate a multi-source collaborative acquisition instruction set corresponding to the target traffic object. The back-injection update module is used to control the corresponding acquisition source to perform adaptive acquisition processing on the target traffic object based on the multi-source collaborative acquisition instruction set, and to back-inject and update the incremental traffic data returned after adaptive acquisition processing, so as to update the unified traffic description fragment set.

10. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the road traffic multi-source information collection method as described in any one of claims 1 to 8.

Images