How Does a License Plate Recognition System Work?

1. Vehicle detection

As the vehicle approaches the entrance, a loop detector or photocell triggers detection; the camera starts capturing images on this signal and focuses on the vehicle's plate area.

2. Plate reading

ANPR/OCR software analyzes the characters in the plate image and converts them to text; image preprocessing steps are applied to improve accuracy for dirty, angled, or nighttime captures.

3. Authorization check

The read plate is compared within milliseconds against whitelist (authorized), blacklist (banned), and guest list records to decide whether the vehicle has transit authorization.

4. Barrier command

For an authorized vehicle, an open command is sent to the barrier or gate control unit via relay, TCP/IP, or API; for an unauthorized vehicle, the barrier stays closed and an optional alarm is triggered.

5. Reporting

The transit is logged in the central database along with the plate, date, time, direction, and vehicle image. These records can be queried for entry-exit reports, occupancy tracking, or security review.

Quick technical facts

What happens inside the camera? From image to text

“Plate reading” is not a single operation but several stages running one after another inside the camera or software. The success of license plate recognition systems depends on each stage working correctly:

1. Locating the plate: The image is searched for a rectangular area resembling a plate; the car body, advertising text and signs are filtered out.
2. Image enhancement: The tilt of the plate area is corrected and contrast and brightness are balanced; IR images are used at night.
3. Character segmentation: The letters and digits on the plate are separated one by one; screws, frames and dirt are distinguished from characters.
4. Character recognition (OCR): Each character is converted into a letter or digit and the plate text is formed.
5. Format validation: Results that do not fit the Turkish plate structure (province code 01–81, letter group, digit group) are discarded; confusions such as “0/O” and “8/B” are reduced at this step.

The SOLEsys Dahua PRO camera we use performs these operations on its own processor; alongside the plate it can also distinguish vehicle type (10 classes), make (147 logos) and colour (12 colours).

Where is the decision made? Camera, site computer and cloud

Whether a read plate is authorized can be checked in three different places. Which one is chosen defines the system type:

A setup that waits for every decision from the cloud can stop traffic when the internet goes down. That is why in Sole systems the on-site device always decides; see our offline license plate recognition guide for details.

How does the system treat different vehicles?

• Subscriber / resident vehicle: The plate is on the list, so the barrier opens without waiting and the passage is recorded with a photo.
• Guest vehicle: The manager or resident defines the plate for a set period; access is free during that time and ends automatically when it expires.
• Unknown vehicle: The barrier stays closed; the operator sees the plate and photo on screen and may grant manual passage, which is also recorded.
• Blacklisted vehicle: Access is blocked and the operator is alerted.
• Paid parking customer: The entry time is recorded by plate and the duration is calculated at exit; the barrier opens after payment at the booth, kiosk, via QR code or at the exit terminal. Details: parking fee by plate.

Safety layer: loop detector and photocell

Plate recognition decides to open the barrier; safe closing is ensured by the ground loop detector and the photocell. The loop detector prevents the arm from coming down before the vehicle has fully passed under the barrier, and the photocell stops the arm when a pedestrian or vehicle breaks the beam. A system installed without these two components carries a risk of vehicle damage even if the plate is read correctly. Details: barrier loop detector and barrier photocell connection.

The time from reading the plate to the barrier being fully open is largely determined by the opening speed of the barrier motor; recognition is a very small part of it. This is covered on our barrier opening time page.

Installation decisions that affect how the system works

The same working principle gives very different results depending on a few decisions made on site: the angle and distance from which the camera sees the plate, night illumination, how much the vehicle slows at the gate and who updates the lists. These decisions are settled during the free survey. For the camera side see our camera selection guide, for read success see accuracy rate factors, and for all system types and cost logic see the license plate recognition system guide.