Car Plate Recognition:
Car plate recognition, also called license plate recogniton or number plate recognition, it is a technology that uses artificial intelligence and deep learning algorithms to automatically detect and recognize the characters of vehicle license plates.
Car Plate Recognition is a valuable tool in many industries. It is used to identify and track vehicles by capturing their license plate numbers through digital imaging technology. LPRs are widely used for law enforcement, security, fleet management, road access control and tolling applications.
What affects the accuracy of car plate recognition?
Affecting the accuracy of license plate identification includes both internal and external factors.There are two internal factors:
1. Recognition method
There are two different recognition methods in the license plate recognition system, soft recognition and hard recognition. Soft recognition is the algorithm of license plate recognition on computer software, also called background recognition. Hard recognition is an algorithm on the front-end camera, also called front-end recognition.
2. Camera resolution
The higher the resolution of a captured image, the better chances of obtaining accurate recognition results. Low-resolution images can lead to the misreading of license plates and thus inaccurate recognition rate.
There are many external factors that affect the accuracy of license plate recognition, mainly as follows:
1. Illumination:
strong light, low light, backlight, thunderstorm weather and other factors will also affect the license plate recognition rate. Adding fill light can solve this problem.
2. Camera angle
It is important that the camera capturing the license plate be positioned at an appropriate angle for good accuracy. If not, skewed images or unrecognizable license numbers could cause a decrease in recognition rate.
3. Vehicle speed
The camera for license plate recognition will select one of the continuously played pictures as the output picture. Then select different frame numbers as recognition pictures. If the speed of the car is too fast, it is easy to drop the frame, and it will not be recognized. It is recommended to add a speed bump before the barrier.
4. License plate
No matter how advanced the license plate recognition algorithm is, it is impossible to correctly identify license plates with occlusions or large stains. Keeping the license plate clean and unobstructed is the first step to ensure correct recognition.
Car Plate Recognition Camera Parameters:
|
Model No.
|
MS-LPR004
|
|
Support Country
|
Taiwan, Brazil, Singapore, Vietnam, Malaysia, Thailand, Indonesia, India, Hong Kong, Macau, Chile, Colombia, South Africa, Dubai, Azerbaijan, Ireland, Brazil, Bulgaria, Poland, Denmark, Germany, France, Netherlands, Kyrgyzstan, Croatia, Kenya, Romania, Malta,Moldova, Switzerland, Serbia, Slovakia, Turkey,etc.
|
|
Sensor
|
3 million pixel starlight CMOS
|
|
Resolution
|
2560*1440
|
|
Recognition Rate
|
99.5%
|
|
Anti-counterfeiting rate
|
99%
|
|
Recognition angle
|
left and right up to 65°, up and down up to 60°
|
|
Recognition distance
|
2 meters ~ 12 meters
|
|
Speed
|
45km/h
|
|
License Plate Recognition Features
|
Number, Color, Type, Width
|
|
License plate white list
|
Support accurate and intelligent fuzzy matching white list license plate rules
|
|
Smart Calibration
|
Support precise or wildcard smart calibration of license plate number, license plate type and color
|
|
Application
|
Indoor and outdoor
|
|
Working Temperature
|
-30℃~+75℃
|
|
Fill light
|
Yes
|
|
Interface
|
4 alarm inputs, 1 RS485, audio input and output, fill light control, USB and other rich interfaces, that meet requirement ono the road barrier, ground sense coil, LED display, PC, voice intercom terminal, etc
|
|
Voice broadcast
|
English, Chinese (support customized)
|
|
Size
|
443mm*146mm*105mm
|
Car Plate Recognition Camera Features:
◆ 4 Million star-level image effects, dual-core processor CNN acceleration;
◆ Equipped with VIR6.0 algorithm and deep learning technology. supports comprehensive and accurate identification of vehicle;
◆ MS-LPR003 can identify 11 types of cars, SUVs, trucks and etc. At the same time support more than 2100 car models;
◆Millisecond-level extremely fast response:From recognition to barrier opening, it will be reduced to 100ms that improve user experience;
◆ Latest anti-counterfeiting algorithm will integrate the characteristics of vehicles and license plates to avoid misidentification on car plate number, alphabet and images;
◆ Cloud management platform breaks through the geographical limitations of traditional networks;
◆ Designed for unmanned, easy installation and maintenance;
◆ Remote voice intercom: The camera has a built-in audio input/output interface;
◆ 4G/5G cableless networking: The stability and reliable on unmanned scenarios determine the operational efficiency and user experience;
◆ Rich hardware interface, stable and reliable performance；
◆ Support TF card data storage, to store number plate identification content, incoming car video and other data, with system logs；
