Licence plate recognition (LPR, also known as ANPR) is a vehicle access control system that reads an arriving vehicle's number plate with a camera, checks it against allow and deny lists, and automatically raises the barrier for registered vehicles. Every passage is archived with plate, date, time and photo, and security staff are alerted when a vehicle is not recognised.
Success depends less on the camera brand than on camera angle, night-time illumination, vehicle detection triggering and how the lists are managed — and by whom. For the wider set-up covering staff, visitor and supplier vehicles, see our vehicle access control system guide.
What actually happens at the gate
At the vehicle entrance of a factory or a residential compound, a security guard stops every car, writes the plate in a logbook or types it into a screen, and phones someone when unsure. Queues build at the morning shift change and again in the evening; records go missing at night and during handovers. After an incident, answering "when did that vehicle enter and leave?" turns into hours of searching logbook pages and camera footage. For residential sites, we cover vehicle entry together with resident and guest flows in residential complex access control.
At logistics hubs the picture is even clearer. Hundreds of lorries and vans come and go every day, and which loading bay each vehicle goes to is decided at the gate. Every minute lost there becomes a team waiting on the dock and a delayed shipment.
The cost of manual vehicle records
Manual vehicle logging costs three things: time lost at the gate, incomplete records and missing evidence after an incident. The best public data on the scale of automated plate reading comes from the United Kingdom:
According to the National ANPR Service data protection impact assessment published on gov.uk, 12,076 camera sets and 1,878 mobile ANPR cameras across the country submit on average more than 100 million plate "read" records to national ANPR systems every day.
The same document states that this data is stored for one year. Taken together, these figures carry a clear lesson: plate reading generates data at very large volumes, and who keeps it, for how long and for what purpose must be defined from the outset. The same question applies at a factory gate.
Organisations increasingly rely on connected devices to secure their premises. Eurostat's "Use of Internet of Things in enterprises" shows that in 2021, 72% of EU enterprises using IoT used it to secure their premises with cameras and similar devices. And MarketsandMarkets forecasts that the global access control market will grow from USD 10.62 billion in 2025 to USD 15.80 billion by 2030. The vehicle gate is often the last entrance to be brought into that transformation.
How licence plate recognition works
- The vehicle is detected. An inductive loop in the road surface or a radar tells the system that a vehicle has reached the reading point. Correct triggering lets the camera capture the sharpest frame.
- The camera captures the plate. An LPR camera with infrared illumination produces a readable image at night as well. Camera angle, distance and shutter speed are decisive here, and a housing exposed to rain and dust needs a suitable IP65/IP66 rating.
- The software reads the plate. The plate region is located in the image and the characters are read. A model tuned to the Turkish plate format correctly separates the province code, letter group and number group, and is trained to read dirty or faded characters. This step rests on the same computer vision foundation as the systems that reject faulty parts on a production line, described in our guide to computer vision quality inspection.
- The lists are checked. The plate is looked up in the allow and deny lists.
- The decision is applied. The barrier opens for a registered vehicle; for an unknown or blocked vehicle, security staff receive an alert.
- The passage is archived. Plate, date, time and vehicle photo are stored and can be searched later by plate or date.
What determines read accuracy?
The same camera can perform very differently at two different gates. These are the factors we see most often on site:
| Factor | Problem | Remedy |
|---|---|---|
| Camera angle and distance | Plate appears skewed or small; characters merge | Correct height and angle per lane |
| Night and oncoming headlights | Plate is too dark or washed out by glare | LPR camera with infrared illumination |
| Vehicle speed | Motion blur | Slow-down zone, correct shutter speed and triggering |
| Dirty or damaged plates | Missing or wrong characters | Reading model trained on these conditions |
| Triggering | Camera captures the vehicle too early or too late | Vehicle detection by loop or radar |
| List management | A correctly read plate is still refused | Authorised staff updating the lists from the panel |
The last row is often overlooked: the system reads the plate correctly, but if a departed employee's car was never removed from the list, or a new contractor's van never added, the decision is still wrong. Plate recognition is not a camera project; it is an authorisation management project. You will find the full component list — camera, loop detector, edge computer and panel — on our licence plate recognition page, and barrier and road blocker options on our barrier systems page.
A falling read rate (a dirty lens, a camera knocked out of alignment, failed infrared illumination) is only noticed if someone is watching it. Organisations running several gates track that indicator alongside camera and barrier status on a remote monitoring platform. Our other guides to cameras, sensors and field devices are gathered in the IoT & Hardware hub.
Data protection: a number plate is personal data
Because a plate can be linked to the vehicle's owner, it counts as personal data under KVKK — Türkiye's Personal Data Protection Law No. 6698, broadly comparable to the GDPR — and so do passage photos. The rules for general area cameras are covered in our guide to CCTV recording and KVKK. Four questions should be answered before installation:
- Purpose: Why is the plate being read — security, parking charges, shipment tracking?
- Notice: Is there a privacy notice at the entrance explaining that recording takes place and why?
- Retention: How long are passage records and photos kept, and what happens when that period ends? The written framework for this is your data retention and disposal policy.
- Access: Who can view the records, and is every view logged?
Settling these questions at the start of the project is far easier than fixing them later; for the wider picture see our KVKK compliance roadmap. A business processing camera and plate records should also check whether it falls under the Board's exemption criteria or must complete VERBIS registration, Türkiye's data controller registry.
How we approach this at Digital Bridge
On licence plate recognition projects we install the camera, the on-site industrial computer (edge) and the barrier connection; the plate-reading software is developed in-house and tuned to the Turkish plate format.
- Gate survey. We inspect entry and exit points, peak-hour vehicle volumes and manoeuvring space in person, and choose the camera position, angle and vehicle detection method accordingly.
- Barrier and equipment. We select barrier and parking systems — boom barriers, high-speed barriers or road blockers — to suit traffic volumes. If you already have a working barrier, we connect it to the system by adding a control board and plate recognition module rather than replacing it.
- Lists and panel. Staff, contractor, visitor and subscriber vehicles are defined in separate lists; authorised staff update them from a web-based panel, which also provides a live passage screen and historical plate search.
- Integration. Through API links to dispatch, warehouse management (WMS) or ERP systems, logistics hubs can assign each vehicle to a loading bay by its plate.
In logistics and transport, for example: at one logistics hub where we installed the system, registered hauliers now enter without barrier staff and new vehicles are flagged to the security lodge; because every passage is recorded by plate, loading bay assignment and vehicle tracking have become faster. We cover duration and charging in car parks in our parking management system guide, and central management of the camera estate in our IP camera and ONVIF article.
Beyond the vehicle gate: the pedestrian side with SmartPass
Plate recognition manages vehicles, but it is people who enter the site. Once a vehicle is through the barrier, the driver, employee or visitor arrives at the door of a building, production area or warehouse. That side is handled by our own product, SmartPass; the basics are explained in our card access control system guide.
Picture a logistics hub. A registered haulier's lorry enters after its plate is read. To reach the dispatch office, the pre-registered driver collects a temporary visitor card from SmartPass at reception; the card's rights are limited to the visit's access group and duration, it stops working at every point once the visit ends, and it is returned on the way out. We compare other visitor entry methods in QR code visitor access.
A warehouse employee taps a 13.56 MHz RFID card at the turnstile: one read opens the door and creates an attendance record, and at lunch in the canteen a meal is deducted if the employee is entitled to one. No separate time-and-attendance device is needed; timesheets follow the shift plan and reports can be exported to payroll.
Access is granted through the groups people belong to: each group is linked to one or more zones, optionally with a weekly time profile, so the dispatch team can enter the Warehouse zone only during working hours. Every rejected read is logged with its reason, such as no zone permission or outside permitted hours. In emergency mode exits are released and entries are closed, while the dashboard shows how many staff, contractors and visitors are inside at that moment. On the data protection side, visitors are not asked for an ID number, a visitor's name and visit purpose are deleted automatically when the retention period ends, and every change is written to the audit log with its previous and new values.
Next step
To see where your own gate stands, spend a week noting peak-hour waiting times, daily vehicle numbers and your vehicle groups (staff, contractors, suppliers, visitors). Then contact us to plan a gate survey; we will settle the camera position, barrier choice and list management on site and follow up with a written proposal.