Method and system for monitoring passing vehicles in traffic hub

Abstract

The invention discloses a traffic hub passing vehicle monitoring method and system, and relates to the technical field of vehicle monitoring, and the method comprises the steps: obtaining a traffic scene image; a target detection algorithm is adopted to identify vehicles in the traffic scene image in real time, and target information is output; storing and processing the target information and the traffic event data based on an open source framework to obtain historical traffic flow data, and predicting a future traffic flow trend based on the historical traffic flow data; performing level-to-level management on authorities of managers based on an access control framework, and allocating data access authorities; and performing correlation analysis on the related factors of the traffic accident by using a data mining algorithm to obtain an accident liability analysis report. Through a YOLOv8 target detection algorithm and a multi-scale feature fusion technology, high-precision vehicle identification is realized, the detection capability of a small target vehicle is improved, and real-time data processing and management in a complex traffic scene are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle monitoring, and in particular to a method and system for monitoring vehicles passing through a transportation hub. Background Art

[0002] In recent years, with the acceleration of urbanization, traffic volume at transportation hubs has continued to grow, placing tremendous pressure on urban traffic management. Traditional traffic monitoring methods are no longer able to meet the demand for efficient and precise management. Vehicle identification suffers from issues such as missing information and inaccurate identification, especially during peak hours and under complex road conditions. Consequently, the public and regulatory authorities have placed higher demands on the completeness, real-time nature, and accuracy of vehicle information collection. In particular, traffic surveys, congestion analysis, and traffic accident handling have significantly increased reliance on key data such as the type, license plate, and driving trajectory of the detected vehicle.

[0003] However, vehicle management at traditional transportation hubs relies on manual guidance, which is inefficient and prone to errors, causing congestion and safety hazards during peak hours. In addition, the intelligent monitoring system lacks real-time analysis capabilities and is unable to accurately identify vehicle information or handle abnormal events.

[0004] Currently, no effective solutions have been proposed for the problems in related technologies. Summary of the Invention

[0005] In response to the problems in the related art, the present invention proposes a method and system for monitoring vehicles passing through a transportation hub to overcome the above-mentioned technical problems existing in the existing related art.

[0006] To this end, the specific technical solutions adopted in the present invention are as follows:

[0007] According to one aspect of the present invention, a method for monitoring vehicles passing through a transportation hub is provided, the method comprising:

[0008] Acquire traffic scene images;

[0009] Adopting target detection algorithm, it can identify vehicles in traffic scene images in real time and output target information;

[0010] Store and process target information and traffic event data based on an open source framework to obtain historical traffic flow data and predict future traffic flow trends based on historical traffic flow data;

[0011] Manage the permissions of managers in a hierarchical manner based on the access control framework and allocate data access permissions;

[0012] Data mining algorithms are used to conduct correlation analysis on relevant factors of traffic accidents and obtain an accident responsibility analysis report.

[0013] Furthermore, obtaining traffic scene images includes:

[0014] Collect traffic scene images through driver-free cameras;

[0015] Based on image sensors, it converts light signals into digital image signals through the photoelectric effect and supports dynamic range adjustment to adapt to imaging requirements in different environments.

[0016] Furthermore, a target detection algorithm is used to identify vehicles in traffic scene images in real time, and the output target information includes:

[0017] Use the YOLOv8 target detection algorithm to identify vehicles in traffic scene images in real time and output target information;

[0018] The YOLOv8 model network structure is optimized through convolutional layer fusion and half-precision inference technology, reducing video memory usage and improving target vehicle detection accuracy.

[0019] Furthermore, target information and traffic event data are stored and processed based on an open source framework to obtain historical traffic flow data. Future traffic flow trends are predicted based on historical traffic flow data, including:

[0020] Build backend services based on open source frameworks and integrate open source relational databases to store target information and perform data management and processing;

[0021] Long short-term memory networks are used to analyze historical traffic flow data, predict future traffic flow trends, and optimize traffic management strategies based on environmental factors.

[0022] Furthermore, based on the access control framework, the management personnel's permissions are hierarchically managed, and data access rights are allocated, including:

[0023] Based on the access control framework, the permissions of administrator terminals are managed in a hierarchical manner, and data access permissions are allocated according to the administrator's rank;

[0024] Abnormal data is screened and retained in real time, triggering the early warning mechanism and pushing it to the management terminal.

[0025] According to another aspect of the present invention, a system for monitoring vehicles passing through a transportation hub is provided, the system comprising:

[0026] Real-time data acquisition module, used to obtain traffic scene images;

[0027] The vehicle detection and recognition module is used to use the target detection algorithm to identify vehicles in traffic scene images in real time and output target information;

[0028] The data processing and analysis module is used to store and process target information and traffic event data based on an open source framework, obtain historical traffic flow data, and predict future traffic flow trends based on historical traffic flow data;

[0029] The authority management and exception response module is used to manage the hierarchical permissions of managers based on the access control framework and allocate data access rights;

[0030] The traffic accident responsibility assessment module is used to use data mining algorithms to conduct correlation analysis on relevant factors of traffic accidents and obtain an accident responsibility analysis report.

[0031] Furthermore, the real-time data acquisition module includes:

[0032] An image acquisition module, used to capture traffic scene images through a driver-free camera;

[0033] The signal processing module is used to convert light signals into digital image signals through the photoelectric effect based on the image sensor, and supports dynamic range adjustment to adapt to imaging requirements in different environments.

[0034] Furthermore, the vehicle detection and identification module includes:

[0035] The target recognition module uses the YOLOv8 target detection algorithm to identify vehicles in traffic scene images in real time and output target information;

[0036] The model optimization module is used to optimize the YOLOv8 model network structure through convolutional layer fusion and half-precision inference technology, reduce video memory usage, and improve target vehicle detection accuracy.

[0037] Furthermore, the data processing and analysis module includes:

[0038] The data management and processing module is used to build back-end services based on an open source framework and integrate an open source relational database to store target information and perform data management and processing;

[0039] The trend prediction and strategy optimization module is used to analyze historical traffic flow data using long-short-term memory networks, predict future traffic flow trends, and optimize traffic management strategies based on environmental factors.

[0040] Furthermore, the permission management and exception response module includes:

[0041] The authority management module is used to manage the authority of administrator terminals in a hierarchical manner based on the access control framework and allocate data access rights according to the administrator's rank;

[0042] The exception handling and early warning module is used to screen and retain abnormal data in real time, trigger the early warning mechanism and push it to the management terminal.

[0043] The beneficial effects of the present invention are:

[0044] 1. This paper uses the YOLOv8 target detection algorithm and multi-scale feature fusion technology to achieve high-precision vehicle recognition and improve the detection capability of small target vehicles. At the same time, combined with MySQL database connection pool technology and SpringBoot caching mechanism, data write throughput is effectively improved and storage space occupation is greatly reduced. The optimized two-level buffer queue design and composite index structure effectively improve data processing efficiency and ensure real-time data processing and management in complex traffic scenarios.

[0045] 2. This invention combines an LSTM algorithm with a data mining model to rapidly determine accident responsibility with high accuracy. A deep learning model trained on historical accident data rapidly correlates and analyzes accident factors, and a spatiotemporal joint index structure supports rapid retrieval of multidimensional data. Furthermore, the use of a USB driver-free camera for hardware deployment simplifies equipment installation and maintenance, reduces hardware costs, and significantly improves the system's affordability and maintainability. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0047] Figure 1 is a flow chart of a method for monitoring vehicles passing through a transportation hub according to an embodiment of the present invention;

[0048] Figure 2 The figure is a principle block diagram of a vehicle monitoring system for a transportation hub according to an embodiment of the present invention.

[0049] In the picture:

[0050] 1. Real-time data acquisition module; 2. Vehicle detection and identification module; 3. Data processing and analysis module; 4. Authority management and abnormal response module; 5. Traffic accident responsibility assessment module. DETAILED DESCRIPTION

[0051] To further illustrate each embodiment, the present invention provides drawings, which are part of the disclosure of the present invention. They are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. By referring to these contents, ordinary technicians in this field should be able to understand other possible implementation methods and the advantages of the present invention.

[0052] According to an embodiment of the present invention, a method and system for monitoring vehicles passing through a transportation hub are provided.

[0053] The present invention will now be further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to the method for monitoring vehicles passing through a transportation hub according to an embodiment of the present invention, the method includes:

[0054] S1. Acquire traffic scene images;

[0055] S2, using target detection algorithms to identify vehicles in traffic scene images in real time and output target information;

[0056] S3: Store and process target information and traffic event data based on an open source framework to obtain historical traffic flow data and predict future traffic flow trends based on historical traffic flow data;

[0057] S4. Manage the permissions of managers in a hierarchical manner based on the access control framework and allocate data access permissions;

[0058] S5. Use data mining algorithms to conduct correlation analysis on relevant factors of traffic accidents and obtain an accident responsibility analysis report.

[0059] Specifically, the data mining algorithm is used to analyze the correlation between traffic accidents and factors such as vehicle type, driver behavior, and road environment, and an accident responsibility analysis report is generated in real time.

[0060] In this optional embodiment, acquiring the traffic scene image includes:

[0061] Collect traffic scene images through a driver-free camera (i.e., a USB driver-free camera);

[0062] Based on image sensors (i.e., CMOS image sensors), light signals are converted into digital image signals through the photoelectric effect, and dynamic range adjustment is supported to adapt to imaging requirements in different environments.

[0063] It should be noted that the USB driver-free camera captures traffic scene images and supports plug-and-play functionality, requiring no additional driver installation. It is highly compatible and adaptable to multiple operating systems (such as Windows and Linux). The image sensor uses CMOS technology, converting light signals into digital image signals through the photoelectric effect, supporting dynamic range adjustment to adapt to imaging needs in different environments.

[0064] In this optional embodiment, a target detection algorithm is used to identify vehicles in traffic scene images in real time, and the output target information includes:

[0065] Use the YOLOv8 target detection algorithm to identify vehicles in traffic scene images in real time and output target information;

[0066] The YOLOv8 model network structure is optimized through convolutional layer fusion and half-precision inference technology, reducing video memory usage and improving target vehicle detection accuracy.

[0067] It should be noted that the YOLOv8 object detection algorithm is used to identify vehicles in images in real time, outputting key information (i.e., target information) such as vehicle type (car, truck, etc.), license plate number, and color. The YOLOv8 model's network structure is optimized, and convolutional layer fusion and half-precision inference technology are used to reduce video memory usage and improve detection accuracy for small targets (such as distant vehicles).

[0068] In this optional embodiment, target information and traffic event data are stored and processed based on an open source framework to obtain historical traffic flow data, and future traffic flow trends are predicted based on the historical traffic flow data, including:

[0069] Build backend services based on an open source framework (i.e., SpringBoot framework) and integrate an open source relational database (i.e., MySQL database) to store target information and perform data management and processing;

[0070] Long short-term memory networks are used to analyze historical traffic flow data, predict future traffic flow trends, and optimize traffic management strategies based on environmental factors.

[0071] It's important to note that the backend service is built using the Spring Boot framework and integrated with a MySQL database to store vehicle information, traffic events, camera parameters, and other data. An LSTM (Long Short-Term Memory) network analyzes historical traffic flow data to predict future traffic flow trends and optimizes traffic management strategies based on weather, time of day, and other environmental factors.

[0072] In this optional embodiment, hierarchical management of administrators' permissions based on the access control framework and allocation of data access permissions include:

[0073] Based on the access control framework (i.e. SpringSecurity framework), the permissions of the administrator's terminal are managed in a hierarchical manner, and data access rights are allocated according to the administrator's rank;

[0074] Abnormal data is screened and retained in real time, triggering an early warning mechanism and pushed to the management terminal (i.e. APP).

[0075] It should be noted that the app implements hierarchical permission management based on the SpringSecurity framework, assigning data access permissions based on job level (e.g., senior managers can view global data, while ordinary employees are limited to regional data). Abnormal data (such as speeding, overloading, and fatigue driving) is screened and retained in real time, triggering an early warning mechanism and pushing it to the manager's terminal.

[0076] According to another embodiment of the present invention, Figure 2 As shown, a traffic hub vehicle monitoring system is also provided, the system comprising:

[0077] Real-time data acquisition module 1, used to acquire traffic scene images;

[0078] Vehicle detection and recognition module 2, used to use target detection algorithm to identify vehicles in traffic scene images in real time and output target information;

[0079] Data processing and analysis module 3, used to store and process target information and traffic event data based on an open source framework, obtain historical traffic flow data, and predict future traffic flow trends based on historical traffic flow data;

[0080] The authority management and exception response module 4 is used to manage the authority of managers in a hierarchical manner based on the access control framework and allocate data access rights;

[0081] The traffic accident responsibility assessment module 5 is used to perform correlation analysis on relevant factors of traffic accidents using a data mining algorithm to obtain an accident responsibility analysis report.

[0082] What needs to be explained is the connection relationship and function of each module:

[0083] Real-time data acquisition module 1 (USB driver-free camera):

[0084] Function: Collect traffic scene images, support plug-and-play, and adapt to multiple operating systems.

[0085] Connection relationship: Directly connected to the main control device (such as PC or server) through the USB interface, the image data is transmitted to the vehicle detection and identification module 2 in real time.

[0086] Vehicle detection and identification module 2 (RTSP protocol):

[0087] Function: Efficiently transmits video streams to the data processing and analysis module 3 through the RTSP protocol, supporting dynamic bandwidth allocation and data retransmission mechanisms.

[0088] Connection relationship: After receiving the camera video stream, it is transmitted to the data processing and analysis module 3 via the network (Ethernet or WiFi).

[0089] Data processing and analysis module 3 (YOLOv8 algorithm):

[0090] Function: Perform real-time target detection on video streams and output information such as vehicle type, license plate number, and location.

[0091] Connection relationship: Receive video data from the vehicle detection and identification module 2, store the results in the MySQL database after processing, and push them to the authority management and exception response module 4 (APP).

[0092] Permission management and exception response module 4 (APP):

[0093] Function: Provides functions such as hierarchical permission management, abnormal event warning, data visualization, and supports multi-terminal access (mobile phone / PC).

[0094] Connection relationship: Interact with the data processing and analysis module 3 and the traffic accident responsibility assessment module 5 through the RESTful API to obtain vehicle information, accident analysis reports, etc. in real time.

[0095] Traffic Accident Responsibility Assessment Module 5 (MySQL):

[0096] Function: Stores structured data such as vehicle information, traffic events, and camera parameters, supporting high-concurrency reading and writing.

[0097] Connection relationship: Bidirectional connection with data processing and analysis module 3 and authority management and exception response module 4, providing data persistence and query services.

[0098] Specifically, the YOLOv8 algorithm refers to the eighth-generation target detection model based on the YOLO framework, and the optimized version is used for vehicle recognition.

[0099] Specifically, RTSP protocol: refers to the Real-Time Streaming Protocol, which is used for efficient transmission of video streams.

[0100] Specifically, MySQL database refers to the open source relational database used by the system to store and manage structured data.

[0101] What needs to be explained is:

[0102] 1) High-precision vehicle detection and identification:

[0103] Advantages: The system uses the YOLOv8 target detection algorithm, and through convolutional layer fusion and multi-scale feature extraction technology, it achieves a vehicle detection accuracy of up to 98% and improves the small target recognition capability by 12%.

[0104] source:

[0105] The optimization of the YOLOv8 algorithm (convolutional layer fusion, half-precision inference) enhances the model's adaptability to complex traffic scenarios.

[0106] The multi-scale feature fusion strategy (backbone network + neck network) improves the detection capability of small target vehicles at a long distance.

[0107] 2) Efficient data processing and storage:

[0108] Advantages: Through MySQL database connection pool technology and SpringBoot caching mechanism, data write throughput reaches 550 records / second and storage space usage is reduced by 60%.

[0109] source:

[0110] The two-level buffer queue design (memory queue + batch storage) optimizes data storage efficiency.

[0111] MySQL composite index structure (timestamp, coordinate, frame number) improves query response speed (reduced by 76%).

[0112] 3) Real-time accident responsibility assessment:

[0113] Advantages: Combining the LSTM algorithm and data mining model, the time for accident responsibility determination is shortened to less than 3 minutes, with an accuracy rate exceeding 95%.

[0114] source:

[0115] The deep learning model (association rule mining and cluster analysis) trained based on historical accident data enables rapid association analysis of accident factors.

[0116] The spatiotemporal joint index structure supports fast retrieval of multidimensional data and shortens the preparation time of model input data.

[0117] 4) Low-cost hardware deployment:

[0118] Advantages: Use USB driver-free camera, plug and play, no additional driver installation is required, and hardware deployment costs are reduced by 50%.

[0119] source:

[0120] The USB camera has strong compatibility and is compatible with multiple operating systems (Windows / Linux), significantly reducing the complexity of hardware maintenance.

[0121] In this optional embodiment, the real-time data acquisition module 1 includes:

[0122] An image acquisition module, used to capture traffic scene images through a driver-free camera;

[0123] The signal processing module is used to convert light signals into digital image signals through the photoelectric effect based on the image sensor, and supports dynamic range adjustment to adapt to imaging requirements in different environments.

[0124] In this optional embodiment, the vehicle detection and identification module 2 includes:

[0125] The target recognition module uses the YOLOv8 target detection algorithm to identify vehicles in traffic scene images in real time and output target information;

[0126] The model optimization module is used to optimize the YOLOv8 model network structure through convolutional layer fusion and half-precision inference technology, reduce video memory usage, and improve target vehicle detection accuracy.

[0127] In this optional embodiment, the data processing and analysis module 3 includes:

[0128] The data management and processing module is used to build back-end services based on an open source framework and integrate an open source relational database to store target information and perform data management and processing;

[0129] The trend prediction and strategy optimization module is used to analyze historical traffic flow data using long-short-term memory networks, predict future traffic flow trends, and optimize traffic management strategies based on environmental factors.

[0130] In this optional embodiment, the rights management and exception response module 4 includes:

[0131] The authority management module is used to manage the authority of administrator terminals in a hierarchical manner based on the access control framework and allocate data access rights according to the administrator's rank;

[0132] The exception handling and early warning module is used to screen and retain abnormal data in real time, trigger the early warning mechanism and push it to the management terminal.

[0133] In summary, with the aid of the above technical solutions of the present invention, the present invention realizes high-precision vehicle recognition through the YOLOv8 target detection algorithm and multi-scale feature fusion technology, significantly improving the detection capability of small target vehicles. At the same time, combined with the MySQL database connection pool technology and the SpringBoot cache mechanism, the data write throughput is effectively improved, the storage space occupied is greatly reduced, and the optimized two-level buffer queue design and composite index structure effectively improve the data processing efficiency, ensuring real-time data processing and management in complex traffic scenarios. The present invention combines the LSTM algorithm and the data mining model to quickly determine the responsibility for the accident and achieve high-accuracy results. Through the deep learning model trained based on historical accident data, the accident factors are quickly correlated and analyzed, and the spatiotemporal joint index structure supports rapid retrieval of multi-dimensional data. In addition, the use of USB driver-free cameras for hardware deployment simplifies the equipment installation and maintenance process, reduces hardware costs, and significantly improves the economy and maintainability of the system.

[0134] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for monitoring vehicles passing through a transportation hub, characterized in that: The method includes: Acquire traffic scene images; Adopting target detection algorithm, it can identify vehicles in traffic scene images in real time and output target information; Store and process target information and traffic event data based on an open source framework to obtain historical traffic flow data and predict future traffic flow trends based on historical traffic flow data; Manage the permissions of managers in a hierarchical manner based on the access control framework and allocate data access permissions; Data mining algorithms are used to conduct correlation analysis on relevant factors of traffic accidents and obtain an accident responsibility analysis report.

2. A method for monitoring vehicles passing through a transportation hub according to claim 1, characterized in that: Acquiring a traffic scene image includes: Collect traffic scene images through driver-free cameras; Based on image sensors, it converts light signals into digital image signals through the photoelectric effect and supports dynamic range adjustment to adapt to imaging requirements in different environments.

3. A method for monitoring vehicles passing through a transportation hub according to claim 1, characterized in that: The target detection algorithm is used to identify vehicles in traffic scene images in real time, and the output target information includes: Use the YOLOv8 target detection algorithm to identify vehicles in traffic scene images in real time and output target information; The YOLOv8 model network structure is optimized through convolutional layer fusion and half-precision inference technology, reducing video memory usage and improving target vehicle detection accuracy.

4. A method for monitoring vehicles passing through a transportation hub according to claim 1, characterized in that: The target information and traffic event data are stored and processed based on the open source framework to obtain historical traffic flow data, and the future traffic flow trend is predicted based on the historical traffic flow data, including: Build backend services based on open source frameworks and integrate open source relational databases to store target information and perform data management and processing; Long short-term memory networks are used to analyze historical traffic flow data, predict future traffic flow trends, and optimize traffic management strategies based on environmental factors.

5. The method for monitoring vehicles passing through a transportation hub according to claim 1, characterized in that: The hierarchical management of management personnel's permissions based on the access control framework and the allocation of data access permissions include: Based on the access control framework, the permissions of administrator terminals are managed in a hierarchical manner, and data access permissions are allocated according to the administrator's rank; Abnormal data is screened and retained in real time, triggering the early warning mechanism and pushing it to the management terminal.

6. A system for monitoring vehicles passing through a transportation hub, used to implement the method for monitoring vehicles passing through a transportation hub according to any one of claims 1 to 5, characterized in that: The system includes: Real-time data acquisition module, used to obtain traffic scene images; The vehicle detection and recognition module is used to use the target detection algorithm to identify vehicles in traffic scene images in real time and output target information; The data processing and analysis module is used to store and process target information and traffic event data based on an open source framework, obtain historical traffic flow data, and predict future traffic flow trends based on historical traffic flow data; The authority management and exception response module is used to manage the hierarchical permissions of managers based on the access control framework and allocate data access rights; The traffic accident responsibility assessment module is used to use data mining algorithms to conduct correlation analysis on relevant factors of traffic accidents and obtain an accident responsibility analysis report.

7. A vehicle monitoring system for a transportation hub according to claim 6, characterized in that: The real-time data acquisition module includes: An image acquisition module, used to capture traffic scene images through a driver-free camera; The signal processing module is used to convert light signals into digital image signals through the photoelectric effect based on the image sensor, and supports dynamic range adjustment to adapt to imaging requirements in different environments.

8. A vehicle monitoring system for a transportation hub according to claim 6, characterized in that: The vehicle detection and identification module includes: The target recognition module uses the YOLOv8 target detection algorithm to identify vehicles in traffic scene images in real time and output target information; The model optimization module is used to optimize the YOLOv8 model network structure through convolutional layer fusion and half-precision inference technology, reduce video memory usage, and improve target vehicle detection accuracy.

9. A vehicle monitoring system for a transportation hub according to claim 6, characterized in that: The data processing and analysis module includes: The data management and processing module is used to build back-end services based on an open source framework and integrate an open source relational database to store target information and perform data management and processing; The trend prediction and strategy optimization module is used to analyze historical traffic flow data using long-short-term memory networks, predict future traffic flow trends, and optimize traffic management strategies based on environmental factors.

10. A vehicle monitoring system for a transportation hub according to claim 6, characterized in that: The rights management and exception response module includes: The authority management module is used to manage the authority of administrator terminals in a hierarchical manner based on the access control framework and allocate data access rights according to the administrator's rank; The exception handling and early warning module is used to screen and retain abnormal data in real time, trigger the early warning mechanism and push it to the management terminal.

Images