A barrier control method and system

Abstract

A barrier control method and system, comprising the following steps: S1. obtaining video streams of a plurality of barriers; S2. performing frame extraction processing on the obtained video streams to obtain to-be-processed frames; S3. performing polling mode license plate recognition on the to-be-processed frames according to the time sequence of the to-be-processed frames, and performing recognition feedback on the barriers where the license plate recognition is successful. The application only obtains video streams on the barriers, then centrally processes the video streams of the barriers, and in the processing process, instead of according to the acquisition time of the video streams, the to-be-processed frames obtained based on the video streams are polled, that is, the to-be-processed frames are processed according to the acquisition time, which can effectively avoid the problem of being stuck in a certain video stream due to poor quality of a certain video stream, reduce the overall construction cost, and still not affect the passing efficiency of the entire barrier.

Description

Technical Field

[0001] This application relates to a barrier gate control method and system. Background Technology

[0002] With the acceleration of urbanization and the continuous growth of motor vehicle ownership, parking lot management is facing increasing pressure. Traditional parking management systems typically adopt an architecture of "AI cameras + backend servers or high-end industrial control computers," which has gradually revealed the following shortcomings in long-term application practice: First, the system architecture is complex and the deployment cost is high.

[0003] In existing technologies, a common parking management system employs terminal processing. This involves hardware decoding of the video stream data captured by cameras at the terminal to complete license plate recognition; the data is then uploaded to a backend server for billing calculations and gate control. This architecture requires cameras with powerful computing and storage capabilities, resulting in high hardware costs. Furthermore, if the parking lot is large with numerous entrances and exits, a large amount of communication cabling is required, leading to long construction periods and high wiring costs. Another common solution is to use an industrial control computer for each lane and individual gate as the billing calculation and gate control module. However, the high cost of industrial control computers further increases the overall system cost.

[0004] Second, the response latency is high and the real-time performance is insufficient.

[0005] In existing systems, the process from camera capture to the final gate opening often involves multiple transmission links: "camera → switch → cloud server → switch → front-end display screen (payment QR code) → third-party payment platform → cloud server → switch → gate controller." Data is repeatedly transmitted between these devices, and server-side queue backlogs contribute to high end-to-end latency. Especially during peak hours, multiple concurrent requests can further exacerbate latency, impacting traffic flow. These multi-level transmission links also consume network bandwidth and server resources, resulting in unnecessary resource waste.

[0006] Third, the multi-gate management capability is insufficient and the scalability is poor.

[0007] When a parking lot has multiple entrances and exits (multiple gate barriers), existing systems typically require each gate barrier to be equipped with an independent controller or industrial computer, with each gate barrier operating independently. To enhance unified management, separate servers are needed, which not only increases the number of devices and deployment costs but also makes maintenance more difficult.

[0008] IV. The layout of indoor parking lots with slopes and curves is limited. In indoor parking lots, especially multi-story parking lots, the space is limited, the field of vision of vehicles is limited when going uphill, downhill and turning, the single camera pattern recognition efficiency is poor, and the driver may even need to make multiple adjustments to the direction. There are also more difficult vehicle operations such as starting on a slope and making small-radius adjustments, which require a high level of driver skill. Alternatively, a special area may be set up with a barrier gate, which occupies parking spaces.

[0009] V. Complex offline payment process Heavy reliance on the network means that if the front-end equipment loses its connection to the cloud server for some reason, the customer payment process becomes complicated and difficult, which can easily cause congestion at the entrance and exit.

[0010] In view of the shortcomings of the existing technologies, there is an urgent need to provide a smart parking management system and method that can reduce system costs, reduce network dependence, achieve low-latency real-time communication, support complex scenario deployment, and support unified management of multiple gates. Summary of the Invention

[0011] To address the aforementioned issues, this application proposes a barrier gate control method, comprising the following steps: S1. Obtaining video streams from several barrier gates; S2. Performing frame extraction processing on the acquired video streams to obtain frames to be processed; S3. Based on the acquisition time of the frames to be processed, performing license plate recognition in a polling manner on the frames to be processed, and sending recognition feedback to the barrier gate that successfully recognizes the license plate. This application only acquires video streams from the barrier gate and then processes them centrally. Furthermore, during processing, it does not rely on the acquisition time of the video streams, but rather polls the frames to be processed obtained from the video streams, i.e., processing is performed according to the acquisition time of the frames to be processed. This effectively avoids the problem of being stuck on a particular video stream due to poor quality, reducing overall construction and installation costs while maintaining the overall passage efficiency of the barrier gate.

[0012] Preferably, the video stream is acquired as follows: a vehicle detection trigger connected to an inductive loop and / or radar sensor and / or infrared sensor detects whether a vehicle has arrived at the gate entrance or exit; when a vehicle arrival signal is detected, the camera is activated to acquire a video stream of a set duration; when the vehicle detection trigger does not detect a vehicle arrival, the camera is controlled to enter a low-power standby state. This application activates the camera for capture and recognition only when a vehicle arrival signal is detected by the vehicle detection trigger. During idle periods without vehicle traffic, the camera is in a low-power standby state or operates at a reduced frame rate, effectively reducing the overall system power consumption and conforming to the trend of green energy conservation.

[0013] Preferably, the video stream is acquired in the following manner: a vehicle detection trigger connected to a ground induction coil and / or radar sensor and / or infrared sensor detects whether a vehicle has arrived at the gate entrance or gate exit; when a vehicle arrival signal is detected, the camera is activated to acquire a video stream of a set duration; when the vehicle detection trigger does not detect a vehicle arrival, the camera is controlled to be in a low image acquisition frame rate state.

[0014] Preferably, the frame extraction process is performed as follows: the video stream is first hardware decoded, and then images are extracted from the video stream at a rate of 25-30 frames per second to obtain the frames to be processed.

[0015] Preferably, the license plate recognition is performed as follows: recognition is performed once for each frame to be processed or for every N frames to be processed in the same video stream, where N is an integer greater than 1.

[0016] Preferably, the identification feedback is performed as follows: the identified license plate number is compared with the locally stored data to determine the vehicle's access permission, and the fee payable is calculated based on the parking duration and local billing rules; a gate control command is generated based on the access permission and the fee payable, and the gate control command is sent to the gate.

[0017] Preferably, after completing the identification feedback, the identification feedback result is transmitted to the cloud management platform; when offline from the cloud management platform, the license plate matching, fee calculation and gate control are completed independently based on the locally stored data, and the data during the offline period is automatically synchronized after the connection with the cloud management platform is restored.

[0018] Preferably, each barrier gate is assigned an independent identifier. Based on the identifier of the corresponding barrier gate in the identification feedback, a barrier gate control command is issued to the corresponding barrier gate, realizing independent control of multiple barrier gates. In this application, each barrier gate has an independent identifier. The local server module can issue control commands to the corresponding barrier gate based on the barrier gate information carried in the identification result, realizing centralized and unified management and independent control of multiple barrier gates, effectively reducing the number of devices and deployment costs in parking lots with multiple entrances and exits.

[0019] Preferably, one of the gates is equipped with several cameras so that one gate corresponds to several video streams. After acquiring the video stream of the gate and completing the recognition feedback, the number of vehicles to be released is accumulated according to the number of vehicles in the gate. When the vehicle reaches the point where the recognition feedback is completed, if the recognition feedback is passed, it is released directly.

[0020] Another aspect of this application discloses a barrier gate control system, including the following modules: The camera module is used to acquire video streams from several gate barriers; The local server module is used to obtain video streams from the camera module, extract frames from the video streams to obtain frames to be processed, and perform real-time license plate recognition on the frames to be processed according to the order in which they were obtained, and send recognition feedback to the gate when the license plate recognition is successful.

[0021] The cloud management platform is used to receive locally stored data from the local server module and to identify feedback.

[0022] Preferably, the local server module uses an ARM architecture embedded processor and integrates an NPU; The camera module uses a PoE camera, a WiFi camera, or a 5G camera and is directly powered by a local server module. The local server module and the cloud management platform are connected via 4G, 5G, Wi-Fi, or WLAN networks and use a real-time communication protocol such as WebSocket, MQTT, or SSE. The real-time communication between the local server module and the cloud management platform includes: uplink communication, where the local server module pushes license plate recognition result data packets to the cloud management platform, the data packets containing the license plate number, vehicle characteristic information, recognition confidence level, parking duration, fees due, and matching status; downlink communication, where the local server module receives control command data packets from the cloud management platform, the control commands including gate opening commands, gate closing commands, billing rule synchronization commands, and license plate list synchronization commands; and heartbeat keep-alive, where the local server module and the cloud management platform exchange heartbeat messages at preset time intervals, and automatically execute a reconnection mechanism when a communication connection is detected to be broken.

[0023] The control interface between the camera module and the local server module is any one or more of the following: GPIO interface, RS485 interface, RS232 interface, or CAN interface. This application uses the local server module as the core processing unit, centralizing core functions such as license plate recognition, toll calculation, and gate control. The local server module employs an ARM-based embedded processor and integrates an NPU, significantly reducing hardware costs compared to traditional industrial control computers or backend server solutions. Furthermore, the camera module can be directly powered by the local server module via PoE, eliminating the need for additional power cables and simplifying construction and deployment.

[0024] This application can bring the following beneficial effects: 1. This application only acquires video streams on the barrier gate and then processes the video streams centrally. Moreover, during the processing, it does not rely on the acquisition time of the video stream, but instead polls the frames to be processed based on the video stream, that is, it processes them according to the acquisition time of the frames to be processed. This can effectively avoid the problem of being stuck on a certain video stream due to poor quality. While reducing the overall construction and installation costs, it does not affect the overall passage efficiency of the barrier gate.

[0025] 2. This application activates the camera for capturing and recognition only when a vehicle arrival signal is detected by a vehicle detection trigger. During idle periods when no vehicles are passing, the camera is in a low-power standby state or operates at a reduced frame rate, which effectively reduces the overall power consumption of the system and conforms to the development trend of green energy saving.

[0026] 3. Each barrier gate in this application has an independent identifier. The local server module can issue control commands to the corresponding barrier gate based on the barrier gate information carried in the identification result, thereby realizing centralized and unified management and independent control of multiple barrier gates and effectively reducing the number of devices and deployment costs of parking lots with multiple entrances and exits.

[0027] 4. This application uses a local server module as the core processing unit, centralizing core functions such as license plate recognition, toll calculation, and barrier gate control. The local server module employs an ARM-based embedded processor and integrates an NPU, significantly reducing hardware costs compared to traditional industrial control computers or backend server solutions. Furthermore, the camera module can be directly powered by the local server module via PoE, eliminating the need for additional power cables and simplifying construction and deployment. Attached Figure Description

[0028] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a schematic diagram of Embodiment 1 of this application.

[0029] Figure 2 This is a schematic diagram of Embodiment 2 of this application. Detailed Implementation

[0030] To clearly illustrate the technical features of this solution, the following detailed description, in conjunction with the accompanying drawings, will explain this application in detail.

[0031] like Figure 1 As shown, a barrier gate control method includes the following steps: S1. Obtain video streams from several gate barriers; There are two ways to acquire the video stream. One way is to put the camera into standby mode when no vehicle is detected. That is, the video stream is acquired as follows: a vehicle detection trigger connected to a ground induction coil and / or radar sensor and / or infrared sensor detects whether a vehicle has arrived at the entrance or exit of the barrier gate. When a vehicle arrival signal is detected, the camera is activated to acquire a video stream of a set duration. When the vehicle detection trigger does not detect a vehicle arrival, the camera is controlled to be in a low-power standby state.

[0032] Another approach is to reduce the camera's frame rate when no vehicle is detected: the video stream is acquired as follows: a vehicle detection trigger connected to an inductive loop and / or radar sensor and / or infrared sensor detects whether a vehicle has arrived at the gate entrance or exit; when a vehicle arrival signal is detected, the camera is activated to acquire a video stream of a set duration; when the vehicle detection trigger does not detect a vehicle arrival, the camera is controlled to be in a low image acquisition frame rate state.

[0033] S2. Perform frame extraction on the acquired video stream to obtain the frames to be processed; The frame extraction process is performed as follows: the video stream is first hardware decoded, and then images are extracted from the video stream at a rate of 25-30 frames per second to obtain the frames to be processed.

[0034] S3. Based on the order in which the frames to be processed are obtained, perform license plate recognition in a polling manner on the frames to be processed, and send recognition feedback to the gate that has successfully recognized the license plate.

[0035] The license plate recognition is performed as follows: recognition is performed once for each frame to be processed or for every N frames to be processed in the same video stream, where N is an integer greater than 1.

[0036] The identification feedback is performed as follows: the identified license plate number is compared with the locally stored data to determine the vehicle's access permission, and the fee payable is calculated based on the parking duration and local billing rules; a gate control command is generated based on the access permission and the fee payable, and the gate control command is sent to the gate.

[0037] After completing the identification feedback, the identification feedback result is transmitted to the cloud management platform; when offline from the cloud management platform, the license plate matching, fee calculation and gate control are completed independently based on the locally stored data, and the data during the offline period is automatically synchronized after the connection with the cloud management platform is restored. Each barrier gate is assigned an independent identifier. Based on the identifier of the corresponding barrier gate in the identification feedback, a barrier gate control command is issued to the corresponding barrier gate to realize the independent control of multiple barrier gates.

[0038] In addition, in situations such as ramps, turns, etc., a gate is equipped with several cameras so that one gate corresponds to several video streams. After acquiring the video stream of the gate and completing the recognition feedback, the number of vehicles to be released is accumulated based on the number of vehicles in the gate. When the vehicle reaches the point where the recognition feedback is completed, if the recognition feedback is passed, it is released directly.

[0039] In practical use, this application adopts a polling method, which actually processes the frames to be processed obtained from the video stream of each barrier in real time without any delay. When there are clear conditions such as rain or obstruction, a new video stream will be generated when the vehicle is in a identifiable position, and then a new frame to be processed will be obtained. Therefore, the processing efficiency is higher.

[0040] With a three-lane layout, each lane allowing one-way entry and exit, the construction cost is approximately 15,000 yuan. If a guard booth is included, an additional computer will be installed, costing an additional 5,000 yuan, bringing the total cost to over 50,000 yuan. To improve traffic efficiency, a tidal flow lane can be added, requiring an additional vehicle license plate recognition system, costing approximately 5,000 yuan. The current solution's construction cost is approximately 55,000 yuan. However, using this technology to achieve the above requirements, four lane gates cost approximately 15,000 yuan, one microserver approximately 8,000 yuan, four standard PoE cameras approximately 1,000 yuan, plus cables, poles, and other auxiliary materials, as well as construction costs of 1,000 yuan, bringing the total construction cost to approximately 28,000 yuan. This represents a significant cost reduction of over 40%; the more lanes at the same entrance / exit, the more pronounced the cost advantage.

[0041] Without adding other cameras, in a three-lane test scenario, 30 vehicles simultaneously entered and exited each lane with zero payment. In three tests, the deployment method of this application took a total of 289s, 301s, and 308s, with an average of 299.33s. Using the conventional deployment method, the total time was 293s, 298s, and 310s, with an average of 300.33s, which can be considered to be the same.

[0042] Adding cameras (standard PoE bullet cameras) to each lane at high-traffic entrances and exits allows for early identification of vehicle license plates, enabling faster passage and improving traffic efficiency by over 30%. It also reduces mechanical wear on lane gates, extending the lifespan of mechanical equipment and lowering maintenance costs. In a specific test, a standard PoE bullet camera was placed at the third parking space in each lane, and the test took 199s, 184s, and 201s respectively, with an average of 194.67s.

[0043] like Figure 2 As shown, a barrier gate control system includes the following modules: Camera module 1 is used to acquire video streams from several gate barriers; Local server module 2 is used to obtain video streams from camera modules, extract frames from the video streams to obtain frames to be processed, and perform real-time license plate recognition on the frames to be processed according to the order in which the frames to be processed are obtained, and send recognition feedback to the gate that successfully recognizes the license plate.

[0044] The cloud management platform 3 is used to receive locally stored data from the local server module and to identify feedback.

[0045] The local server module uses an ARM architecture embedded processor and integrates an NPU; The camera module uses a PoE camera, a WiFi camera, or a 5G camera and is directly powered by a local server module. The local server module and the cloud management platform are connected via 4G, 5G, Wi-Fi, or WLAN networks and use a real-time communication protocol such as WebSocket, MQTT, or SSE. The real-time communication between the local server module and the cloud management platform includes: uplink communication, where the local server module pushes license plate recognition result data packets to the cloud management platform, the data packets containing the license plate number, vehicle characteristic information, recognition confidence level, parking duration, fees due, and matching status; downlink communication, where the local server module receives control command data packets from the cloud management platform, the control commands including gate opening commands, gate closing commands, billing rule synchronization commands, and license plate list synchronization commands; and heartbeat keep-alive, where the local server module and the cloud management platform exchange heartbeat messages at preset time intervals, and automatically execute a reconnection mechanism when a communication connection is detected to be broken. The control interface between the camera module and the local server module is any one or more of the following: GPIO interface, RS485 interface, RS232 interface, or CAN interface.

[0046] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0047] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for controlling a barrier gate, characterized in that: Includes the following steps: S1. Obtain video streams from several gate barriers; S2. Perform frame extraction on the acquired video stream to obtain the frames to be processed; S3. Based on the order in which the frames to be processed are obtained, perform license plate recognition in a polling manner on the frames to be processed, and send recognition feedback to the gate that has successfully recognized the license plate.

2. The barrier gate control method according to claim 1, characterized in that: The video stream is acquired as follows: a vehicle detection trigger connected to a ground induction coil and / or radar sensor and / or infrared sensor detects whether a vehicle has arrived at the entrance or exit of the barrier gate; when a vehicle arrival signal is detected, the camera is activated to acquire a video stream of a set duration. When the vehicle detection trigger does not detect the arrival of a vehicle, the camera is controlled to enter a low-power standby state.

3. The barrier gate control method according to claim 1, characterized in that: The video stream is acquired as follows: a vehicle detection trigger connected to a ground induction coil and / or radar sensor and / or infrared sensor detects whether a vehicle has arrived at the entrance or exit of the barrier gate; when a vehicle arrival signal is detected, the camera is activated to acquire a video stream of a set duration. When the vehicle detection trigger does not detect the arrival of a vehicle, the camera is controlled to be in a low image acquisition frame rate state.

4. The barrier gate control method according to claim 1, characterized in that: The frame extraction process is performed as follows: the video stream is first hardware decoded, and then images are extracted from the video stream at a rate of 25-30 frames per second to obtain the frames to be processed.

5. The barrier gate control method according to claim 1, characterized in that: The license plate recognition is performed as follows: recognition is performed once for each frame to be processed or for every N frames to be processed in the same video stream, where N is an integer greater than 1.

6. The barrier gate control method according to claim 1, characterized in that: The identification feedback is performed as follows: the identified license plate number is compared with the locally stored data to determine the vehicle's access permission, and the fee payable is calculated based on the parking duration and local billing rules; a gate control command is generated based on the access permission and the fee payable, and the gate control command is sent to the gate.

7. A barrier gate control method according to claim 6, characterized in that: After completing the identification feedback, the identification feedback result is transmitted to the cloud management platform; when offline from the cloud management platform, the license plate matching, fee calculation and gate control are completed independently based on the locally stored data, and the data during the offline period is automatically synchronized after the connection with the cloud management platform is restored. Each barrier gate is assigned an independent identifier. Based on the identifier of the corresponding barrier gate in the identification feedback, a barrier gate control command is issued to the corresponding barrier gate to realize the independent control of multiple barrier gates.

8. The barrier gate control method according to claim 1, characterized in that: A gate is equipped with several cameras, so that one gate corresponds to several video streams. After the video stream of the gate is acquired and the recognition feedback is completed, the number of vehicles to be released is accumulated according to the number of vehicles in the gate. When the recognition feedback is completed, if the recognition feedback is passed, the vehicle is released directly.

9. A barrier gate control system for implementing any one of the barrier gate control methods described in claims 1-8, characterized in that: Includes the following modules: The camera module is used to acquire video streams from several gate barriers; The local server module is used to obtain the video stream from the camera module and perform frame extraction on the video stream to obtain the frame to be processed. Based on the order in which the frames to be processed are obtained, the system performs real-time license plate recognition on the frames to be processed and sends recognition feedback to the gate that successfully recognizes the license plate. The cloud management platform is used to receive locally stored data from the local server module and to identify feedback.

10. A barrier gate control system according to claim 9, characterized in that: The local server module uses an ARM architecture embedded processor and integrates an NPU; The camera module uses a PoE camera, a WiFi camera, or a 5G camera and is directly powered by a local server module. The local server module and cloud management platform adopt 4G network connection, 5G network connection, WIFI connection or WLAN network connection and real-time communication protocol of WebSocket protocol, MQTT protocol or SSE protocol. The real-time communication between the local server module and the cloud management platform includes: uplink communication, where the local server module pushes license plate recognition result data packets to the cloud management platform, the data packets containing license plate number, vehicle feature information, recognition confidence, parking duration, fees due, and matching status; downlink communication, where the local server module receives control command data packets from the cloud management platform, the control commands including gate opening command, gate closing command, billing rule synchronization command, and license plate list synchronization command; and heartbeat keep-alive, where the local server module and the cloud management platform exchange heartbeat messages at preset time intervals, and automatically execute a reconnection mechanism when a communication connection is detected to be broken. The control interface between the camera module and the local server module is any one or more of the following: GPIO interface, RS485 interface, RS232 interface, or CAN interface.

Images