Using vehicle identification technology to build a visual intelligent parking lot integrated management system

Traditional parking lots have many problems, such as low efficiency, difficult to find parking spaces, difficult to find cars, difficult to manage and high management costs, which seriously restrict the modern development of urban transportation. Convenient, safe, efficient and scientific parking lot integrated management system has become the new demand of parking construction, and visual and intelligent parking lot integrated management system has emerged.

Parking space detection system based on video

1 detection steps

(1) The high-definition camera host acquires the real-time video stream.

(2) Using background subtraction algorithm to detect the moving foreground. Firstly, the background model is established by self-learning the initial multi-frame video images, and then the current frame image and the background model are differentiated to eliminate the influence of the background, so as to obtain the foreground area of the moving target.

(3) According to the detection results of moving objects in the background subtraction operation, the background model is selectively updated and saved.

(4) Filter the noise to obtain the accurate vehicle position.

(5) Using the algorithms of time and space information, matching and prediction, the vehicle is accurately tracked, the trajectory of the vehicle object is obtained, and the trajectory information of the vehicle object is saved.

And (6) judging whether the vehicle reaches the position of the trigger line, if not, detecting the next frame, and if so, sending a trigger signal.

Pure video license plate recognition scene

The gateway management system adopts the gateway management mode of video recognition. At the entrance and exit of the ground/underground parking lot, the vehicle information is collected and identified by the front-end capture camera, and the vehicle information data is sent to the back-end management center by using the network. Vehicle identification technology is used to compare the license plate number, logo, vehicle type, body color and other related data, so as to ensure that vehicles can enter and leave the parking space in a well-documented, controllable and reasonable way.

High-definition entrance capture camera: realize video monitoring, vehicle license plate, logo, body color, model and other related information collection and identification functions;

Vehicle detector: receiving the feedback signal from the ground loop, detecting whether there is a vehicle or not, and feeding back the detection information to realize the functions of triggering snapshot and preventing the vehicle from smashing;

Triggering the ground loop: sensing the vehicle, feeding back the signal to the vehicle detector, and the vehicle detector outputs the trigger signal, triggering the high-definition entrance capture camera to capture, so as to realize the functions of one car, one map, door opening and the like;

Anti-smashing loop: sensing the vehicle, feeding back the signal to the vehicle detector, and the vehicle detector outputs the anti-smashing signal to realize the anti-smashing function of the road gate;

LED display screen: release information such as charging amount, license plate number and free parking space; Displays the number of free parking spaces in the released parking lot.

Self-service payment machine: realize self-service payment function;

Mobile management terminal: realize mobile charging management and support mobile charging, offline charging and license plate photo recognition;

Gateway management terminal: realizes gateway system management control and provides related gateway management application services, and has built-in charging client components;

Cloud platform: provides mobile APP application services, and supports mobile APP application functions such as remaining position release, mobile payment, parking space reservation, reverse car search, parking space guidance, parking lot navigation, etc.

In order to effectively improve the congestion caused by parking/card picking and card swiping at entrances and exits, license plate number identification authentication is used instead of traditional card swiping authentication, which can effectively improve the traffic efficiency of entrances and exits.

Admission can be achieved without stopping at the entrance. Vehicles with fixed exits, vehicles that meet the free time, vehicles that pay their own fees and vehicles that pay fees at the center service desk can leave without stopping, and vehicles that pay fees at the exit kiosk can leave quickly without swiping their cards.

System characteristics

1 has high recognition rate and strong adaptability

(1) High-performance entrance capture camera: 2.5m-6m multi-distance capture, which can adapt to different capture scenes, and the recognition rate is over 99%.

(2) Dual-camera capture recognition: It solves the problem of low license plate recognition rate in non-standard lanes such as ultra-wide lanes and turning lanes, and greatly improves the license plate recognition rate.

(3) Cloud secondary license plate correction and recognition: The wrong license plate can be corrected by cloud secondary recognition, which greatly improves the license plate recognition rate and ensures the normal entry and exit of vehicles.

(4) Handheld terminal recognizes the license plate: In the case of camera failure or network interruption at the entrance and exit of the front door, the license plate can be recognized through the handheld terminal to ensure the reliability and stability of license plate recognition.

(5) Manual check and correction of license plate: For the wrong license plate, the system supports manual check and correction of license plate to ensure the normal entry and exit of vehicles.

(6) Fuzzy matching of license plates: fuzzy matching of fixed vehicles without stopping, and fuzzy matching of temporary vehicles or unlicensed vehicles.

(7) Double detection of license plate capture: The capture camera supports video-triggered capture and coil-triggered capture to run at the same time, which minimizes the phenomenon of missing shots.

2 Visual intelligent management

(1) Real-time monitoring: Real-time monitoring of vehicles entering and leaving the site, so that vehicles and people inside can see clearly, and things can be checked after being controlled.

(2) Direct release of special vehicles: direct release of vehicles with special vehicle identification.

(3) Corresponding release of vehicles and parking spaces: When there are multiple parking spaces in the parking lot, for the owner with exclusive parking spaces, the vehicles must correspond to the parking spaces in a parking lot before they can be released automatically.

(4) Restrict temporary users from entering during peak hours: During peak hours, the parking lot can set corresponding time periods to restrict temporary users from entering.

(5) One car per person: The owner owns multiple cars, but only has one parking space of his own in the venue. When any car of the owner comes into the field, other cars will not be able to come into the field.

(6) Preventing repeated entry and exit: Since the license plate number of the vehicle is the only sign for entering and leaving the parking lot, it is impossible to have repeated entry and exit problems when using the card collection method, which can effectively prevent the occurrence of fee evasion.

(7) No entry without a full seat: temporary carriages can be set up or no entry is allowed when all vehicles are full.

(8) Parent-child parking lot management: When there are multiple parking lots in the parking lot, the system supports the setting of parent-child parking lots, and relevant entry rules can be set for each parking lot according to the specific requirements of specific users. For example, temporarily parked vehicles cannot enter the underground fixed owner parking lot of the sub-hospital after entering through the entrance and exit of the parent hospital.

(9) Anti-blocking of drainage at entrances and exits: By deploying anti-congestion identification cameras, identify the congestion at entrances and exits, and prompt and guide vehicles to enter and leave vacant entrances and exits.

(10) alarm reminder: the system can provide all-round alarm or prompt output such as sound effect, LED and software prompt in real time for all kinds of alarm events such as blacklisted vehicles and full seats.

3 High stability

(1) Off-line work: The system supports off-line work, which can ensure the normal access traffic of vehicles in unexpected situations such as network interruption or management platform software failure. When the network returns to normal, offline access and exit related data will be automatically uploaded to the management platform synchronously.

(2) Dual-machine hot standby: the system supports dual-machine hot standby of management equipment to ensure the stability of the system in all directions.

end