A visual traffic starting point analysis system

Abstract

The present invention relates to a visual traffic starting point analysis system, which is connected to a first image acquisition and analysis unit and a second image acquisition and analysis unit, wherein: the first image acquisition and analysis unit is used to collect traffic flow information at a road intersection, and the second image acquisition and analysis unit is used to collect and determine vehicle information in the traffic flow information that does not comply with a first predetermined rule. The first predetermined rule is called by the analysis system from a data processing platform. The visual traffic starting point analysis system uses the vehicle information that does not comply with the first predetermined rule determined by the second image acquisition and analysis unit as a benchmark, and determines the traffic starting and ending point path information of vehicles that do not comply with the first predetermined rule within at least a part of the time interval by analyzing the traffic flow information of the first image acquisition and analysis unit within the at least part of the time interval.

Description

Technical Field

[0001] The present invention relates to the field of intelligent transportation technology, and in particular to a visual traffic starting point analysis system. Background Art

[0002] In recent years, with the emergence of various forms of big data and advancements in related analysis technologies, the focus of urban transportation researchers has gradually shifted from system operational status monitoring to traffic demand analysis (including vehicle traceability). Obtaining information about vehicle travel origins and associated routes is a key component of traffic demand analysis. However, many traditional detection data have information limitations when calculating vehicle travel origins. For example, GPS positioning data often cannot represent all vehicles due to limited sample sizes, and fixed-point vehicle detection data (such as microwave vehicle detection data) lacks vehicle identity information.

[0003] High-definition intelligent toll-check systems automatically recognize license plate data (hereinafter referred to as toll-check data). The data they capture not only includes vehicle identity information but also corresponds to road sections, eliminating the need for complex matching calculations. In theory, this provides ideal data for calculating vehicle travel starting points. However, in reality, most cities currently struggle to achieve full toll-check coverage across their road networks. Even if toll-check systems are installed at every intersection and road section, vehicle identification errors or omissions often occur due to various factors, such as aging equipment, damaged license plates, and inclement weather. Therefore, calculating the starting points of motor vehicle travel on a road network using incomplete and flawed toll-check data is challenging.

[0004] Patent publication number CN109035784A discloses a method for estimating dynamic traffic starting points based on multi-source heterogeneous data. This method integrates license plate recognition data and vehicle GPS data, splitting road network cross-sectional flow into observable and unobservable traffic starting points. The unobservable traffic starting points are estimated using a Kalman filter and then integrated with the observable traffic starting points to obtain dynamic traffic starting point information. This method combines vehicle data collected by license plate recognition equipment with GPS data to extract valid paths as observable traffic starting point information, which is then divided into observable and unobservable components to estimate road network traffic starting points. This method only supplements the license plate recognition data with vehicle GPS data already entered into the system to fill in and complete the missing parts of the license plate recognition data. However, this method cannot guarantee that the supplemented data effectively covers the entire road network. This method involves multi-faceted analysis of a single dataset, and its application is similarly focused on analyzing a single phenomenon as a means of analyzing traffic dynamics.

[0005] To address the issue of existing internet data being limited by sample size, leading to deviations in confidence and accuracy, the present invention utilizes traffic flow information collected by a first image acquisition and analysis unit and vehicle information collected by a second image acquisition and analysis unit to aggregate data from a full sample size. This eliminates the need to consider constraints imposed by internet road data, such as user numbers and online time, and eliminates the need for new samples, thus reducing the investment cost of sample development. The present invention enables precise analysis of the status of intersections, road sections, and road networks by leveraging full-time data from all vehicles, intersection direction-level traffic data, and lane-level traffic data from sections.

[0006] In addition, on the one hand, there are differences in understanding among those skilled in the art; on the other hand, the inventor studied a large number of documents and patents when making the present invention, but due to space limitations, not all details and contents are listed in detail. However, this does not mean that the present invention does not have the characteristics of these prior arts. On the contrary, the present invention already has all the characteristics of the prior art, and the applicant reserves the right to add relevant prior art to the background art. Summary of the Invention

[0007] To address the shortcomings of the prior art, the present invention provides a visual traffic starting point analysis system. The visual traffic starting point analysis system addresses the fact that traffic flow data currently collected on the internet suffers from sample size limitations, leading to certain deviations in data confidence and accuracy. Therefore, the analysis system integrates a full sample size of data by introducing multi-source vehicle and traffic flow information acquired by different image acquisition units, eliminating the need to consider constraints such as user number and online time on obtaining internet traffic flow data. The visual traffic starting point analysis system is connected to a first image acquisition and analysis unit and a second image acquisition and analysis unit, wherein the first image acquisition and analysis unit is configured to collect traffic flow information at a road intersection, and the second image acquisition and analysis unit is configured to collect and determine vehicle information within the traffic flow information that does not comply with a first predetermined rule. The first predetermined rule is invoked by the visual traffic starting point analysis system from a data processing platform. The visual traffic starting point analysis system uses the vehicle information determined by the second image acquisition and analysis unit as a reference to determine traffic starting and ending point path information for vehicles that do not comply with the first predetermined rule within at least a portion of a time interval by analyzing the traffic flow information collected by the first image acquisition and analysis unit within at least a portion of a time interval. Its advantage is that the present application uses the vehicle information that does not conform to the first predetermined rule determined by the first image acquisition and analysis unit to trace back the traffic starting and ending point path information of the vehicle in the traffic flow information collected by the first image acquisition and analysis unit, thereby utilizing the vehicle information that does not conform to the first predetermined rule and the traffic starting and ending point path information of the vehicle that does not conform to the first predetermined rule generated by the same vehicle being captured successively by the second image acquisition and analysis unit and the first image acquisition and analysis unit within a certain time interval, thereby restoring the driving path and itinerary information of the vehicle based on the data information collected by the second image acquisition and analysis unit and the first image acquisition and analysis unit. The visual traffic starting point analysis system provided by the present application accurately analyzes the status of intersections, road sections, and road networks through the full-time data of all-domain vehicles, intersection direction-level traffic, and section lane-level traffic. In order to avoid the mutual constraints of software and hardware at the computing power level, the present application optimizes and adapts software and hardware through reasonable core components to ensure the effective and stable computing power of the computing environment. At the same time, the compatibility and security of the interface are taken into consideration to facilitate the access and output of multi-source data; in order to avoid algorithm efficiency degradation or mismatch in special scenarios, with the support of full sample data and multi-scenario learning, the algorithm is continuously iterated to achieve multi-scenario replication and application, thereby enhancing the universality of the software.The present application can effectively combine the data collected by the first image acquisition and analysis unit and the data collected by the second image acquisition and analysis unit in the industry intranet, eliminate the singleness of the first image acquisition and analysis unit or the second image acquisition and analysis unit when collecting data, especially the setting areas of the two on the road can compensate each other, so that the import lane-level traffic flow at the road intersection position is combined with the cross-sectional traffic flow in the road section. By mutual verification of the acquisition results of the two, data missing, data anomalies and the like caused by missing vehicle travel can be corrected, thereby constructing complete path information of the vehicle within a certain time interval.

[0008] According to a preferred embodiment, the traffic starting and ending point path information of the vehicle collected by the first image acquisition and analysis unit is replaced in a certain time period, and the time point corresponding to the traffic starting and ending point path information of the vehicle collected by the first image acquisition and analysis unit and the time point of the vehicle information determined by the second image acquisition and analysis unit as not complying with the first predetermined rule are in the same time interval; the data processing platform traces back the traffic starting and ending point path information of the specified vehicle in the traffic flow information collected by the first image acquisition and analysis unit within the same time interval based on the vehicle information determined by the second image acquisition and analysis unit as not complying with the first predetermined rule, thereby determining the travel information of the specified vehicle by splicing the vehicle driving path information of the vehicle information confirmed by the second image acquisition and analysis unit and the traffic starting and ending point path information of the vehicle collected by the first image acquisition and analysis unit. Its advantage lies in that the data processing platform obtains vehicle information through the second image acquisition and analysis unit, and filters out the traffic start and end point information of the vehicle captured by the second image acquisition and analysis unit from the traffic flow information collected by the first image acquisition and analysis unit within the same time interval. The data processing platform then uses the traffic start and end point information of the vehicle at the road intersection and the vehicle's driving information on the lanes in the road network to deduce or splice the vehicle's driving path and travel information within a certain time interval, making it easier to analyze and judge the traffic flow conditions of different roads and road intersections at different time points, thereby facilitating relevant personnel to formulate vehicle driving rules for the road network based on the obtained traffic flow information. By repairing the missing data of vehicle information, more complete and accurate vehicle travel information can be obtained, making it easier for relevant departments and users to make adaptive traffic plan adjustments based on road conditions more accurately.

[0009] According to a preferred embodiment, the vehicle information determined by the second image acquisition and analysis unit that does not comply with the first predetermined rule can be used as traceability information for the data processing platform to track information, so that the data processing platform can filter out the traffic start and end point path information of the specified vehicle that does not comply with the first predetermined rule from the traffic start and end point path information of the traffic flow in the same time interval collected by the first image acquisition and analysis unit.

[0010] According to a preferred embodiment, the second image acquisition and analysis unit calls the reference driving information that meets the first predetermined rule in the data processing platform and compares it with the traffic flow information collected by it, marks the vehicles in the traffic flow information collected by it that do not meet the first predetermined rule, and the data processing platform uses the vehicle information of the marked vehicles to filter out at least part of the travel information of the specified vehicle including the traffic start and end point information from the traffic flow information.

[0011] According to a preferred embodiment, the vehicle information determined by the second image acquisition and analysis unit as not complying with the first predetermined rule refers to driving information of the vehicle when it is traveling outside the driving rules pre-set on the data processing platform and / or driving information of the vehicle collected by the second image acquisition and analysis unit during a time interval in which at least part of the vehicle's travel information is missing;

[0012] The first predetermined rule includes at least a second predetermined rule that limits the vehicle to travel in accordance with a predetermined travel rule and a third predetermined rule that the vehicle has no missing travel information within at least a partial time interval.

[0013] According to a preferred embodiment, the first image acquisition and analysis unit extracts the number and vehicle information of vehicles passing through the road intersection within at least part of a specific time interval by continuously acquiring image information of all vehicles passing through the road intersection, thereby generating traffic flow information for the specific time interval.

[0014] According to a preferred embodiment, the second image acquisition and analysis unit performs real-time verification of vehicles passing through its acquisition area by calling the first predetermined rule in the data processing platform. When a vehicle that does not match the first predetermined rule appears in the acquisition area of ​​the second image acquisition and analysis unit, the second image acquisition and analysis unit collects vehicle information.

[0015] According to a preferred embodiment, the data processing platform determines the path information between the first image acquisition and analysis unit and the second image acquisition and analysis unit by means of vehicle journey planning, thereby defining a time interval in which the acquisition time points of the first image acquisition and analysis unit and the second image acquisition and analysis unit are in the same time period based on the path information;

[0016] When the time point at which the second image acquisition and analysis unit collects vehicle information that does not comply with the first predetermined rule is obtained, the data processing platform determines the time interval in which the first image acquisition and analysis unit can obtain the traffic start and end point path information of the vehicle that does not comply with the first predetermined rule based on the collection time point of the second image acquisition and analysis unit.

[0017] According to a preferred embodiment, the traffic flow information collected by the first image acquisition and analysis unit refers to the number and vehicle information of vehicles passing through the road intersection within a specific time interval. The data processing platform compares the vehicle information that does not comply with the first predetermined rule with the vehicle information collected by the first image acquisition and analysis unit within the same time interval, thereby obtaining the traffic start and end point path information of the vehicles that do not comply with the first predetermined rule.

[0018] According to a preferred embodiment, the vehicle information collected by the second image acquisition and analysis unit that does not comply with the first predetermined rule can supplement the travel information of the vehicle passing through two road intersections continuously, thereby generating complete path information of the vehicle within a certain time interval. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a flow chart of a preferred embodiment of a visual traffic starting point analysis system of the present invention;

[0020] Figure 2 It is a schematic diagram of the architecture of a preferred embodiment of a visual traffic starting point analysis system of the present invention.

[0021] Reference Signs List

[0022] 1: First image acquisition and analysis unit; 2: Second image acquisition and analysis unit; 3: Data transmission unit; 4: Data processing platform; 41: Application layer; 42: Support layer; 43: Interface layer; 44: Data layer; 45: Access layer; 46: Physical layer; 411: Application program; 421: Geographic information module; 422: System management module; 423: Traffic starting point big data analysis module; 431: Basic data system interface; 432: Operation data system interface; 433: Business application system interface; 441: Application support service module; 442: Map support service module; 443: Acquisition database; 444: Shared support database; 445: User database; 446: Clearing and settlement database; 447: Basic information database; 448: Filing database; 449: Subject analysis database; 461: Host facilities; 462: Storage facilities; 463: Network facilities; 464: Security facilities. DETAILED DESCRIPTION

[0023] The following is a detailed description with reference to the accompanying drawings.

[0024] Example 1

[0025] according to Figure 1 A visual traffic starting point analysis system is shown, which may include a first image acquisition and analysis unit 1 , a second image acquisition and analysis unit 2 , a data transmission unit 3 and a data processing platform 4 .

[0026] According to a specific embodiment, the first image acquisition and analysis unit 1 and the second image acquisition and analysis unit 2 upload the collected vehicle information and traffic flow information to the data processing platform 4 through the data transmission unit 3. The data processing platform 4 stores the first predetermined rule that can be used to judge the driving condition of the vehicle in a real-time updated manner. The data transmission unit 3 can retrieve the latest version of the first predetermined rule from the data processing platform 4 to judge the driving condition of the vehicle passing through its collection area. The data processing platform 4 can also use the vehicle information that does not comply with the first predetermined rule in the traffic flow information determined by the data transmission unit 3 to filter the traffic flow information at the road intersection position collected by the first image acquisition and analysis unit 1, so that the data processing platform 4 can obtain the traffic start and end point path information of the vehicle whose driving condition does not comply with the first predetermined rule when passing through the first image acquisition and analysis unit 1 within the same time period. The data processing platform 4 can piece together the complete driving path and itinerary information of the vehicle based on the vehicle information of the designated vehicle that does not comply with the first predetermined rule within the same time period and the traffic start and end point path information of the vehicle. Thus, the data processing platform 4 analyzes the traffic flow formed when vehicles in the entire road network pass through road intersections and road lanes at all times, and accurately obtains the traffic flow status of intersections, road sections, and road networks.

[0027] Preferably, the time points at which the first image acquisition and analysis unit 1 captures the route information corresponding to the specified vehicle's starting and ending points and the time points at which the second image acquisition and analysis unit 2 determines that the specified vehicle's travel does not conform to the first predetermined rule fall within the same time interval. This allows the data processing platform 4 to assemble the vehicle's potential travel path based on the sequential order of the time points recorded by the first and second image acquisition and analysis units 1 and 2, and to determine the specified vehicle's complete travel path and itinerary information based on multiple, dispersed first and second image acquisition and analysis units 1 and 2. Vehicle information that does not conform to the first predetermined rule refers to driving information such as vehicle images, basic vehicle information, and driver images captured by the second image acquisition and analysis unit 2 when the vehicle's travel does not conform to the first predetermined rule. By combining the travel path and itinerary information of all vehicles in the road network over all time periods, the data processing platform 4 can determine traffic flow at intersections and road lanes at different times, thereby facilitating relevant departments and users to more accurately adjust adaptive traffic plans based on road conditions.

[0028] Preferably, the data processing platform 4 uses the travel information of the designated vehicle determined by the second image acquisition and analysis unit 2 as traceability information for tracing the designated vehicle's travel path in the road network. The data processing platform 4 uses the travel information of the designated vehicle to filter out the traffic start and end point path information of the designated vehicle that does not meet the first predetermined rule from the traffic start and end point path information of the traffic flow within the same time interval collected by the first image acquisition and analysis unit 1. Then, the data processing platform 4 determines the travel information of the designated vehicle by splicing the vehicle travel path information using the vehicle information determined by the second image acquisition and analysis unit 2 and the traffic start and end point path information of the vehicle collected by the first image acquisition and analysis unit 1.

[0029] Preferably, the first predetermined rule called by the second image acquisition and analysis unit 2 can be the reference driving information of vehicles that comply with the driving rules pre-loaded by the data processing platform 4, so that the second image acquisition and analysis unit 2 verifies whether the vehicles passing through its image acquisition area appear to be consistent with the reference driving information of vehicles that comply with the driving rules pre-loaded by the data processing platform 4. If the vehicle is captured driving in a manner that does not comply with the first predetermined rule, it is determined that the vehicle does not comply with the first predetermined rule. Preferably, the reference driving information that complies with the first predetermined rule can be driving image data of the vehicle when it is driving on the road in compliance with the pre-set driving rules. When the data processing platform 4 determines that the vehicle is driving in a manner that does not comply with the first predetermined rule, the data processing platform 4 marks the vehicles that do not comply with the first predetermined rule in the traffic flow information it has collected, and the data processing platform 4 uses the vehicle information of the marked vehicles to filter out at least part of the travel information of the specified vehicle including traffic start and end point information from the traffic flow information. Further preferably, the vehicle information determined by the second image acquisition and analysis unit 2 as not complying with the first predetermined rule refers to the driving information of the vehicle when it is traveling in a manner deviating from the driving rules pre-set on the data processing platform 4 and / or the driving information of the vehicle collected by the second image acquisition and analysis unit 2 during a time interval in which at least part of the travel information is missing. Preferably, the first predetermined rule includes at least a second predetermined rule that limits the vehicle to traveling in accordance with the pre-set driving rules and a third predetermined rule that the vehicle does not have any travel information missing within at least part of a specific time interval. Preferably, the specific time interval means: if the vehicle can be collected by the first image acquisition and analysis units 1 at two adjacent road intersections with the traffic start and end point path information while complying with the driving rules, and a single path information is directly constructed, then the time interval between the acquisition time points of the two adjacent first image acquisition and analysis units 1 is the specific time interval.

[0030] Preferably, the first image acquisition and analysis unit 1 continuously collects information about all vehicles passing through the road intersection. When the second image acquisition and analysis unit 2 determines the time of acquisition of a vehicle that does not comply with the first predetermined rule, the data processing platform 4 extracts the number and information of vehicles passing through the road intersection within at least a portion of the time interval captured by the first image acquisition and analysis unit 1 by defining a time interval, thereby generating traffic flow information for the specific time interval. Preferably, the data processing platform 4 determines the driving path of vehicles passing through the first image acquisition and analysis unit 1 and the second image acquisition and analysis unit 2 sequentially through vehicle trip planning, and determines whether the acquisition times of the first image acquisition and analysis unit 1 and the second image acquisition and analysis unit 2 fall within the same time interval based on the actual road conditions along the driving path. When the time of acquiring vehicle information captured by the second image acquisition and analysis unit 2 that does not comply with the first predetermined rule, the data processing platform 4 determines, based on the acquisition time of the second image acquisition and analysis unit 2, the time interval in which the first image acquisition and analysis unit 1 can acquire the traffic origin and destination path information of the vehicle that does not comply with the first predetermined rule.

[0031] like Figure 2 As shown, the data information architecture of the data processing platform 4 includes an application layer 41 that connects it to device terminals, a support layer 42 that provides technical support, an interface layer 43 that facilitates external device access to the data processing platform 4, a data layer 44 that processes data uploaded by the first and second image acquisition and analysis units 1 and 2, an access layer 45 that connects to the data transmission unit 3, and a physical layer 46 that provides the hardware and network infrastructure for the data processing platform 4. Preferably, the support layer 42 manages the exchange and sharing of associated data between the data layer 44 and the application layer 41, including support for data acquisition, data sharing and exchange, data statistical analysis, and geographic information services. The interface layer 43 provides, as needed, a basic data system interface 431, an operational data system interface 432, and a business application system interface 433, allowing users to access and obtain different underlying hardware usage permissions through different interfaces. The data layer 44 stores and analyzes vehicle data uploaded by the access layer 45, utilizing the collected vehicle information for data calculations and analysis, and establishing relevant computational models. The access layer 45 establishes data connections with the first image acquisition and analysis unit 1 and the second image acquisition and analysis unit 2 respectively through the data transmission unit 3 .

[0032] like Figure 2As shown, the application layer 41 is the application 411 loaded onto the terminal device by the traffic origin point analysis system. Preferably, the application 411 may include systems and apps. Preferably, the support layer 42 includes a geographic information module 421, a system management module 422, and a traffic origin point big data analysis module 423. These three modules of the support layer 42 are all centrally managed through data centralization, and the infrastructure of the traffic origin point analysis system is completed by integrating all information into a single, intuitive diagram. Preferably, the interface layer 43 provides an external interface for the data processing platform 4, facilitating administrators' input of control commands. The interface layer 43 includes a basic data system interface 431, an operational data system interface 432, and a business application system interface 433. Preferably, the data layer 44 includes an application support service module 441 and a map support service module 442. Preferably, the application support service module 441 provides the relevant data interfaces required for system display and system function development. Preferably, the map support service module 442 provides GIS services such as map display and planning. Preferably, the access layer 45 is used to input external data such as data collected by the first image acquisition and analysis unit 1 and the second image acquisition and analysis unit 2, map data, and parking lot data into the data processing platform 4, and it can perform data access conversion according to the needs of the data to be accessed. Preferably, the access layer 45 can also be used for user authorization services. Preferably, the physical layer 46 includes host facilities 461, storage facilities 462, network facilities 463 and security facilities 464. Preferably, the host facilities 461 are servers (computing centers) that support computing and application deployment. The storage facilities 462 are storage hardware that supports data storage. The network facilities 463 are the communication infrastructure of this system. The security facilities 464 are hardware and software security facilities that ensure the security of data and applications.

[0033] Further preferably, the data layer 44 may also include a collection database 443, a shared support database 444, a user database 445, a clearing and settlement database 446, a basic information database 447, a filing database 448, and a subject analysis database 449. Preferably, information services for regular users can be provided through WeChat and handheld terminal APPs. The data processing platform 4 uses the data collected by the first image acquisition and analysis unit 1 and the second image acquisition and analysis unit 2 to collect data of the entire sample size, eliminating the need for new data acquisition equipment, reducing construction investment costs, and the data processing platform 4 does not need to consider restrictions on Internet road data access rights such as the number of users and online time. The data processing platform 4 accurately analyzes the status of intersections, road sections, and road networks by collecting full-time data of all vehicles in the entire region, intersection flow-level traffic data, and section lane-level traffic data. In addition, in order to avoid mutual constraints between software and hardware at the computing power level, the present invention achieves optimized adaptation of software and hardware through reasonable planning of core components, ensuring the stability of the computer's computing power. At the same time, the present application also optimizes the compatibility and security of data interfaces to facilitate the access and output of multi-source data. Finally, with the support of full sample data and multi-scenario learning, the present invention achieves multi-scenario replication and application by continuously iterating the algorithm, thereby enhancing the universality of the analysis system.

[0034] Preferably, the present invention utilizes data obtained by existing vehicle data acquisition equipment to enable industry applications, the specific scope of which is as follows:

[0035] Traffic control optimization design (analysis of traffic characteristics, special periods, commuting routes, etc.);

[0036] Traffic organization and management (marking, channelization, one-way traffic, etc. optimization analysis);

[0037] Configuration of transportation service facilities (bus routes, motor vehicle parking areas, non-motor vehicle parking areas, etc.);

[0038] Highway network planning (design of ring roads, expressway entrances and exits, road cross sections, setting up interchanges, feasibility studies of new construction or reconstruction projects, etc.);

[0039] Forecast of long-term traffic volume (attraction volume, occurrence volume, transit volume, etc.);

[0040] National economic evaluation and financial analysis of planning schemes and construction projects.

[0041] The visual traffic starting point analysis system involved in the present invention needs to process data collected by different types of traffic sensors, which are highly specific and complex. The visual traffic starting point analysis system uses multi-source data fusion technology to achieve complementary integration of traffic data from different sensors, accurately reflecting the real-time traffic status of roads. Due to the real-time, sudden, and disordered nature of data, the visual traffic starting point analysis system adopts a distributed processing framework to address issues such as excessive data volume, inconsistent data sources across different time and space, and data disorder. This allows the visual traffic starting point analysis system to comprehensively, accurately, and effectively assess real-time traffic operation status. In response to the massive amount of collected data, the visual traffic starting point analysis system applies statistical methods, case-based reasoning, decision trees, and genetic algorithms, among other big data statistical analysis methods, to identify patterns in traffic data. This provides technical support for intelligent transportation design, helping to alleviate traffic congestion, optimize traffic network operation, and promote the healthy and stable development of transportation. Preferably, the system is constructed using a multi-layer B / S (browser / server) architecture, accessible to users via web browsers, WAP browsers, and apps. Traffic management departments can use the visual traffic origin and destination analysis system to monitor urban traffic conditions in real time, enabling them to make traffic forecasts for specific road sections and formulate appropriate traffic management plans. Ideally, the visual traffic origin and destination analysis system can perform temporal and spatial statistical and comparative analysis of operational indices and traffic flows across the entire road network, regions, business districts, and roads, ultimately generating reports comparing congestion indexes, congestion status, traffic flows, and operational indexes.

[0042] Example 2

[0043] This embodiment is a further improvement of embodiment 1, and repeated contents will not be repeated here.

[0044] The present invention also provides a method for visualizing traffic starting point analysis. This method addresses the problem of incomplete points in the first image acquisition and analysis unit (1) of existing road networks, preventing the effective construction of a comprehensive basic database for analyzing motor vehicle travel starting points. By combining the data collected by the first image acquisition and analysis unit (1) with data collected by the second image acquisition and analysis unit (2), map data, and management data, the present invention can repair missing travel information for specific vehicles and expand the sample size of traffic starting point data.

[0045] Preferably, the visualization traffic starting point analysis method provided by the present invention can solve the defects of the prior art through the methods of secondary patching and iterative expansion. Specifically, "secondary patching" refers to the initial patching of the missing vehicle travel information in the traffic flow information obtained by the first image acquisition and analysis unit 1 based on the spatiotemporal behavior characteristics of the vehicle; and secondary patching of the travel information of the vehicle passing through the road section without the first image acquisition and analysis unit 1 based on the vehicle information collected by the second image acquisition and analysis unit 2. Through the above two patches, the distribution of the vehicle's traffic starting points will be closer to the actual situation. "Iterative expansion" refers to extracting the vehicle traffic starting points from the data collected by the first image acquisition and analysis unit 1, and allocating the vehicle's initial traffic starting points to the road network, and then using the road section flow as the total amount constraint, performing multiple iterations on the vehicle traffic starting points collected by the first image acquisition and analysis unit 1 to achieve sample expansion. Preferably, the iterative expansion analysis process can make the number of different types of traffic starting points closer to the actual situation.

[0046] Preferably, a visual traffic starting point analysis method may include the following steps:

[0047] S1: Point information verification and preprocessing.

[0048] S11: Verification and preprocessing of static point information.

[0049] (1) Point verification. In actual single data information collection, the point information of the first image acquisition and analysis unit 1 may have problems such as coordinate errors, incorrect or ambiguous names, and inaccurate equipment time. This will affect the accuracy of the subsequent information sorting, input and analysis by the traffic flow big data integration module and the judgment and analysis module. Therefore, in order to ensure the accurate calculation of the monitored travel information, the present invention needs to verify the static data collected by the first image acquisition and analysis unit 1 and the captured data collected by the second image acquisition and analysis unit 2, so that the present invention can check and modify the above problems one by one by mutual verification of dynamic and static data.

[0050] (2) Point classification and marking. Preferably, when classifying vehicle information, in order to accurately determine the vehicle's driving position, the first image acquisition and analysis unit 1 needs to be classified and marked according to the nature of the road section where it is located.

[0051] S12: Dynamic data verification.

[0052] When the first image acquisition and analysis unit 1 monitors traffic flow, data loss often occurs. Therefore, algorithmic patching or sampling expansion is necessary to compensate for the data gaps captured by the first image acquisition and analysis unit 1 and capture the complete vehicle journey information. In theory, a vehicle passing through the first image acquisition and analysis unit 1 located at the upstream mainline entrance will inevitably be detected by the first image acquisition and analysis unit 1 located at the downstream off-ramp exit or the first image acquisition and analysis unit 1 located at the downstream mainline exit. In actual monitoring, only 80%-85% of vehicles passing through the first image acquisition and analysis unit 1 located at the upstream mainline entrance will effectively match a vehicle detected by the first image acquisition and analysis unit 1 located at the downstream off-ramp exit or the first image acquisition and analysis unit 1 located at the downstream mainline exit.

[0053] S2: Trip integrity analysis and initial trip repair.

[0054] Preferably, a complete trip of a vehicle on a road system is defined as a trip in which the license plate recognition information is left when the vehicle enters the ramp or plane entrance (i.e., point O) on the road, and leaves the off-ramp or plane exit (i.e., point D). Any missing information section at point O or point D is called an incomplete trip. The purpose of trip integrity analysis is to classify and count the missing information of the vehicle's trip, so as to adopt different data repair methods for different missing types. Preferably, there is a certain topological relationship between multiple adjacent first image acquisition and analysis units 1 on the same road, and there is also a certain regularity in the driving time and driving path of many vehicles on the road network. By fully exploring, summarizing and refining the above-mentioned rules and characteristics, the vehicle trip information can be repaired (i.e., the initial repair).

[0055] There are several ways to infer missing itinerary information:

[0056] S3: Expansion and secondary repair of traffic starting point.

[0057] S31: Path allocation at the traffic starting point.

[0058] Since the road section flow corresponds to a specific road, the vehicle traffic starting point collected by the first image acquisition and analysis unit 1 must be allocated to a specific road before being expanded, thereby establishing a corresponding relationship between vehicle travel information and the road.

[0059] S32: Expansion of traffic starting point.

[0060] The expansion of traffic starting points is accomplished in two steps. The first is to calculate the ratio of the road section flow rate to the allocated flow rate at the road intersection, known as the cross-sectional flow rate correction factor. The second is to calculate the expansion factor for each type of traffic starting point based on the cross-sectional flow rate correction factor on each road.

[0061] (1) Section flow correction coefficient

[0062] Several sections (called check sections) are evenly selected on the expressway network to check the accuracy of the flow distribution.

[0063] For the check section i, the section flow correction coefficient (denoted as ki) is defined as follows:

[0064] K i =Q i / q i (1)

[0065] where Q i ,q i They are the detected flow rate of the section and the flow rate allocated from the traffic starting point.

[0066] Based on these characteristics, a phased sampling strategy was adopted. Specifically, the central urban area, where the first image acquisition and analysis units (1) are more prevalent and trips are relatively complete, was processed separately from the peripheral areas, where the first image acquisition and analysis units (1) are relatively less prevalent. Prioritizing the sampling of road sections in the central urban area, the problem was concentrated in the periphery. Subsequently, various vehicle trajectory data (including surveillance equipment capture data, taxi GPS, and user-generated content (UGC) trajectory data) were used to centrally address missing trip directions in the peripheral sections. Finally, a unified, iterative sampling of trips at all traffic origins was performed to ensure that the estimated traffic origin distribution was as close to the actual situation as possible.

[0067] (2) Traffic starting point expansion coefficient

[0068] Considering that high-flow calibration sections have a greater impact on the overall system deviation, the present invention uses a proportional weighted method to calculate the final sample expansion coefficient (denoted as K). That is, the K value of a certain type of traffic starting point is equal to the weighted average of the cross-sectional flow correction coefficients of each road it passes through according to its detected flow value. Therefore, the number of samples after the expansion of this traffic starting point is:

[0069] OD'=OD×K (2)

[0070] (3) Evaluation criteria for sample expansion effect

[0071] In practice, the aforementioned "allocation - expansion - verification - redistribution - further expansion" process is repeated repeatedly. After each round of expansion, different traffic origins are enlarged or reduced by varying degrees. To measure the overall expansion effect, the root mean square error (RMSE) of the difference between the allocated and measured flows at all verification sections is used as an evaluation criterion. If the RMSE value decreases with each expansion, the expansion results are improving.

[0072]

[0073] where q i 'Distribute the traffic flow at the starting point of the verification section i after sample expansion, Q i is the detection flow rate within the section, and m is the number of verification sections.

[0074] S32: Secondary repair of travel information.

[0075] In peripheral areas, due to the relatively small number of points in the first image acquisition and analysis unit 1, some vehicles' travel information is inevitably completely missing, resulting in distorted distribution of traffic origins. This problem cannot be solved by simply expanding the sample size of existing traffic origins. Relying on other data sources to supplement the missing travel information is necessary to provide a more realistic representation of vehicle travel information. GPS data offers the advantages of clear and widespread trajectory data. Therefore, the present invention utilizes GPS trajectory data to perform a secondary repair of travel information in peripheral areas where the first image acquisition and analysis unit 1 is severely missing.

[0076] Preferably, the GPS trajectory data in the present invention can also be replaced by Beidou trajectory data collected by Beidou satellites.

[0077] Example 3

[0078] The present invention provides an application of a visual traffic starting point analysis method. This method leverages data acquired by existing acquisition equipment and systems (a first image acquisition and analysis unit 1 and a second image acquisition and analysis unit 2) to enable industry applications. This method can be applied to projects such as road control signal optimization design, traffic organization and management, highway network planning, and feasibility studies for new or renovated projects. The present invention can provide a quantitative basis for predicting long-term traffic volumes, determining road types and grades, establishing interchanges, designing road cross sections, configuring traffic service facilities, managing and controlling traffic, and conducting national economic evaluations and financial analyses of construction projects, thereby laying the foundation for improved traffic planning and scientific decision-making for construction projects.

[0079] Preferably, the traffic starting point analysis system utilizes multi-data fusion technology and big data mining and analysis technology to mine and analyze the data collected by the access layer 45, thereby generating information that can be viewed by urban traffic management departments and other departments, and helping users to manage and make scientific decisions, thereby improving the service level of urban traffic. This allows the traffic starting point analysis system to effectively enhance the data value and application efficiency compared to existing technologies. Preferably, the visual traffic starting point analysis method can include a traffic condition monitoring unit, a research and analysis unit, a data reporting unit, a traffic flow big data integration unit, an urban parking management function integration unit, a geographic information unit, and a system management unit. The method includes the following:

[0080] 1. Traffic Condition Monitoring Unit: This unit categorizes and integrates monitoring data, performs correlation analysis based on GIS road network data, and provides real-time information on urban road traffic conditions. It can be used to view data on traffic starting points at intersections, road sections, and regions, and can also view real-time monitoring images of intersections through a video interface.

[0081] 2. Research and Analysis Unit: Analyze the traffic volume at intersections within the jurisdiction. It can also conduct qualitative and quantitative research and analysis on data such as average vehicle speed, travel time, and traffic flow trends.

[0082] 3. Data reporting unit: It includes node intersection statistics and feature reports, commuting route statistics and feature reports, transit route statistics and feature reports, holiday route statistics and feature reports, important unit perimeter route statistics and feature reports, district and jurisdiction boundary road traffic feature reports, custom range traffic feature reports, loop entrance and exit and diversion feature analysis, micro-circulation road excavation and analysis, and parking resource and demand analysis.

[0083] 4. Traffic flow big data integration unit: Access, store, extract, clean, correlate, compare and analyze traffic data such as traffic flow and video through unified interface specifications and adaptive access services.

[0084] 5. Urban parking management function integration unit: supports functions such as parking lot location information entry and review, layer annotation and query display; supports intuitive viewing of traffic conditions and videos of roads around the parking lot on the map; supports query of parking lot location and attributes; supports query of the number of parking spaces and the number of vacant parking spaces in the parking lot; supports display of parking space status, and can use different colors to mark whether the parking space is vacant; supports query of license plates of vehicles parked in parking spaces in the parking lot; supports query of parking space vacancy rate.

[0085] 6. Geographic information unit: supports the query and search of geographic location information, and renders the collected real-time geographic location information data on the map.

[0086] 7. System Management Unit: Monitors, warns, and audits the operating status of the system and the command platform's internal modules. It authenticates the legitimacy of users logging into the system, manages and controls permissions for users and their roles, and records platform function operations and user operations in real time, allowing for query and statistics on log information.

[0087] It should be noted that the above-mentioned specific embodiments are exemplary, and those skilled in the art can come up with various solutions inspired by the disclosure of the present invention, and these solutions also fall within the scope of the disclosure of the present invention and fall within the scope of protection of the present invention. Those skilled in the art should understand that the description of the present invention and its drawings are illustrative and do not constitute a limitation on the claims. The scope of protection of the present invention is defined by the claims and their equivalents. Throughout the text, the features guided by "preferably" are only an optional method and should not be understood as having to be set, so the applicant reserves the right to abandon or delete the relevant preferred features at any time.

Claims

1. A visual traffic starting point analysis system, comprising a first image acquisition and analysis unit (1) and a second image acquisition and analysis unit (2), wherein: The first image acquisition and analysis unit (1) is used to acquire traffic flow information at a road intersection. The second image acquisition and analysis unit (2) is used to acquire and determine vehicle information in the vehicle flow information that does not comply with a first predetermined rule, wherein the first predetermined rule is called by the visual traffic starting point analysis system from the data processing platform (4), It is characterized by: The visual traffic starting point analysis system is based on the vehicle information determined by the second image acquisition and analysis unit (2) that does not conform to the first predetermined rule, and determines the traffic starting and ending point path information of the vehicle that does not conform to the first predetermined rule within at least a part of the time interval by analyzing the traffic flow information of the first image acquisition and analysis unit (1) within at least a part of the time interval. The time point corresponding to the vehicle's traffic start and end point path information collected by the first image acquisition and analysis unit (1) and the time point of the vehicle information not meeting the first predetermined rule determined by the second image acquisition and analysis unit (2) are within the same time interval, The data processing platform (4) traces back the traffic start and end point path information of the designated vehicle in the traffic flow information collected by the first image acquisition and analysis unit (1) within the same time interval based on the vehicle information determined by the second image acquisition and analysis unit (2) that does not conform to the first predetermined rule, thereby determining the travel information of the designated vehicle by splicing the vehicle travel path information of the vehicle information determined by the second image acquisition and analysis unit (2) and the traffic start and end point path information of the vehicle collected by the first image acquisition and analysis unit (1).

2. The visual traffic starting point analysis system according to claim 1, characterized in that: The vehicle information determined by the second image acquisition and analysis unit (2) as not conforming to the first predetermined rule can be used as traceability information for the data processing platform (4) to track information, so that the data processing platform (4) can filter out the traffic start and end point path information of the designated vehicle that does not conform to the first predetermined rule from the traffic start and end point path information of the traffic flow within the same time interval acquired by the first image acquisition and analysis unit (1).

3. The visual traffic starting point analysis system according to claim 2, characterized in that: The second image acquisition and analysis unit (2) calls the reference driving information that meets the first predetermined rule in the data processing platform (4) and compares it with the traffic flow information collected by it, marks the vehicles that do not meet the first predetermined rule in the traffic flow information collected by it, and the data processing platform (4) uses the vehicle information of the marked vehicles to filter out at least part of the travel information of the designated vehicle including traffic start and end point information from the traffic flow information.

4. The visual traffic starting point analysis system according to claim 3, characterized in that: The vehicle information determined by the second image acquisition and analysis unit (2) not conforming to the first predetermined rule refers to driving information of the vehicle when it is traveling in a manner that deviates from the driving rules pre-set on the data processing platform (4) and / or driving information of the vehicle collected by the second image acquisition and analysis unit (2) during a time interval in which at least part of the travel information is missing; The first predetermined rule includes at least a second predetermined rule that limits the vehicle to travel in accordance with a predetermined travel rule and a third predetermined rule that the vehicle has no missing travel information within at least a certain time interval.

5. The visual traffic starting point analysis system according to claim 4, characterized in that: The first image acquisition and analysis unit (1) extracts the number and vehicle information of vehicles passing through the road intersection within at least a part of the time interval by continuously acquiring image information of all vehicles passing through the road intersection, thereby generating traffic flow information for a specific time interval.

6. The visual traffic starting point analysis system according to claim 5, characterized in that: The second image acquisition and analysis unit (2) verifies vehicles passing through its acquisition area in real time by calling a first predetermined rule in the data processing platform (4); when a vehicle that does not match the first predetermined rule appears in the acquisition area of ​​the second image acquisition and analysis unit (2), the second image acquisition and analysis unit (2) acquires vehicle information.

7. The visual traffic starting point analysis system according to claim 6, characterized in that: The data processing platform (4) determines the path information between the first image acquisition and analysis unit (1) and the second image acquisition and analysis unit (2) by means of vehicle journey planning, thereby defining a time interval in which the acquisition time points of the first image acquisition and analysis unit (1) and the second image acquisition and analysis unit (2) are in the same time period based on the path information; When the time point at which the second image acquisition and analysis unit (2) acquires vehicle information that does not conform to the first predetermined rule is obtained, the data processing platform (4) determines, based on the acquisition time point of the second image acquisition and analysis unit (2), a time interval in which the first image acquisition and analysis unit (1) can acquire traffic start and end point path information of the vehicle that does not conform to the first predetermined rule.

8. The visual traffic starting point analysis system according to claim 7, characterized in that: The traffic flow information collected by the first image acquisition and analysis unit (1) refers to the number and vehicle information of vehicles passing through a road intersection within a specific time interval. The data processing platform (4) compares the vehicle information that does not conform to the first predetermined rule with the vehicle information collected by the first image acquisition and analysis unit (1) within the same time interval, thereby obtaining traffic start and end point path information of the vehicles that do not conform to the first predetermined rule.

9. The visual traffic starting point analysis system according to claim 8, characterized in that: The vehicle information collected by the second image acquisition and analysis unit (2) that does not conform to the first predetermined rule can supplement the travel information of the vehicle passing through two road intersections continuously, thereby generating complete path information of the vehicle within a certain time interval.

Images