What are ANPR Cameras?
A guide to Automatic Number Plate Recognition technology
If you have ever driven into a car park that opened automatically, received a toll invoice in the mail, or noticed cameras mounted above motorway gantries, you have likely already had an encounter with ANPR technology. Automatic Number Plate Recognition — commonly abbreviated as ANPR, and sometimes referred to as ALPR (Automatic License Plate Recognition) in North American contexts — is one of the most widely deployed surveillance and enforcement technologies in the modern world.
Put simply, ANPR cameras automatically read and record vehicle registration plates. In the seconds it takes a car to pass a camera, the system captures an image, extracts the plate text, and cross-references it against databases — all without any human intervention. From law enforcement to private car park management, the applications are broad, and the technology is advancing rapidly.
For a comprehensive technical overview, see the Wikipedia article on Automatic Number Plate Recognition.
How Do ANPR Cameras Work?
At its core, ANPR is a combination of specialised cameras, image processing software, and optical character recognition (OCR). The process happens in a rapid sequence of steps, typically completing in around 250 milliseconds — faster than a blink of an eye.
Step 1: Image Capture
The process begins the moment a vehicle enters the camera’s field of view. ANPR cameras are purpose-built for this task — they typically use high-speed shutters to freeze motion and avoid blur, even when vehicles are travelling at highway speeds. Many systems incorporate infrared (IR) illumination, which floods the number plate with near-infrared light invisible to the human eye. This allows the camera to capture a crisp, high-contrast image of the plate regardless of the ambient lighting conditions — whether it is blazing midday sun, the dead of night, or the glare of oncoming headlights.
Number plates are manufactured to be highly reflective to infrared light, which is why they appear almost white or luminous in ANPR camera footage, even when they look quite ordinary to the naked eye.
Step 2: Plate Localisation
Once an image is captured, the software must identify exactly where the number plate is within it. This is called plate localisation. The system analyses the image for rectangular regions of high contrast and specific aspect ratios consistent with number plates. Modern systems use machine learning models trained on millions of plate images to recognise plates even when they are partially obscured, at an angle, or dirty.
Step 3: Image Normalisation & Enhancement
After locating the plate, the software crops, rotates, and scales the plate region to a standardised size and orientation. Enhancement techniques may be applied to sharpen edges, increase contrast, or reduce noise — all to make the subsequent character recognition as accurate as possible.
Step 4: Optical Character Recognition (OCR)
This is where the magic happens. The processed plate image is fed into an OCR engine specifically designed for number plate fonts and layouts. Rather than a generic document OCR tool, ANPR OCR is tuned to the specific character sets, fonts, and spacing used by number plates in the target jurisdiction. For example, when the Netherlands updated its vehicle registration plate font in 2002, small gaps were deliberately introduced into letters like P and R specifically to make them more distinguishable by ANPR systems.
The result is a text string — the number plate — extracted in digital form.
Step 5: Database Lookup & Action
The extracted plate number is then checked against one or more databases in real time. Depending on the use case, this might mean checking:
• A police “hot list” of stolen vehicles, disqualified drivers, or wanted persons
• A whitelist of authorised vehicles for a private car park
• A toll payment database to determine whether a fee is owed
• A parking duration log to determine how long a vehicle has been in a zone
If there is a match — or a mismatch where one is expected — the system triggers the appropriate response: alerting an officer, raising or lowering a barrier, issuing a fine, or simply logging the event for future analysis.
The Technology Behind ANPR
ANPR technology has come a long way since its invention. Here is a look at the key components that make modern systems so powerful.
A Brief History
ANPR was invented in 1976 at the Police Scientific Development Branch in Britain. Prototype systems were operational by 1979, and early deployments appeared on the A1 road and at the Dartford Tunnel. The first arrest facilitated by ANPR — the detection of a stolen car — came in 1981. However, it was not until the 1990s, when cheaper PC-based hardware and more flexible software arrived, that ANPR became practical for widespread use.
A landmark moment came in November 2005 in Bradford, UK, when ANPR played a pivotal role in locating and convicting the killers of police officer Sharon Beshenivsky — one of the first documented cases of ANPR helping to solve a murder.
Camera Hardware
Modern ANPR cameras are compact, weatherproof units that contain the full image capture and often the initial processing hardware within a single housing. Key hardware features include:
• High-resolution image sensors capable of capturing sharp images of fast-moving plates
• Integrated infrared LED illuminators for night-time and low-light operation
• Adjustable focal lengths to accommodate wide lanes, gantry mounting, or close-range car park installations
• Ruggedised enclosures rated for outdoor temperature extremes, rain, and dust
Software & Artificial Intelligence
The software side of ANPR has been transformed by machine learning. Contemporary systems use convolutional neural networks (CNNs) — the same class of AI models used in facial recognition and self-driving vehicles — to perform plate localisation and character recognition. This dramatically improves accuracy in real-world conditions, where plates may be obscured by mud, damaged, poorly lit, or mounted at unusual angles.
Get Corey Hinde’s stories in your inbox
Join Medium for free to get updates from this writer.
The seven primary algorithmic steps an ANPR system typically performs are: plate localisation, plate orientation and sizing, normalisation, character segmentation, optical character recognition, syntactical and geometrical analysis, and finally, results reporting. Each step has been refined over decades to achieve the high accuracy rates modern systems deliver.
Fixed vs. Mobile Systems
ANPR systems come in two main forms. Fixed installations are mounted at permanent locations — toll plazas, car park entrances, motorway gantries, or border crossings. They provide continuous monitoring of a specific point. Mobile systems are mounted on vehicles, typically police patrol cars, and can scan plates of parked or moving vehicles as the patrol car drives past. Mobile systems must handle the additional complexity of relative speeds potentially exceeding 160 km/h when reading oncoming traffic.
Processing: At the Edge or in the Cloud?
Some systems perform all processing locally at the camera itself — this is known as edge processing, and it is preferred where a fast, real-time response is needed, such as raising a car park barrier. Other systems, particularly those managing large volumes of cameras (such as the London Congestion Charge network), transmit images to centralised server farms where OCR is performed in batches. Cloud-based approaches allow for more powerful processing and easier database integration, at the cost of a network dependency.
📹 See It In Action
Want to see how ANPR cameras actually look and operate in real life? This short video from Georges Car Media gives a great visual demonstration of the technology in action:
https://www.tiktok.com/@georgescarmedia/video/7377324182226472225
Who Uses ANPR Cameras?
ANPR technology has found applications across a remarkably wide range of industries and government agencies. Below is a look at some of the major users.
New Zealand Police
New Zealand Police use ANPR cameras extensively, both in fixed installations and mounted on patrol vehicles. The cameras are linked to the NZTA motor vehicle register and police databases, allowing officers to instantly identify vehicles of interest — including stolen cars, vehicles with expired or no registration, and those associated with wanted persons or outstanding warrants.
Mobile ANPR units in patrol cars can passively scan hundreds of plates per hour on a normal patrol route, without the officer needing to manually check each one. This dramatically increases the chances of identifying offenders who might otherwise go undetected among normal traffic.
New Zealand Transport Agency (NZTA / Waka Kotahi)
NZTA uses ANPR as part of its traffic monitoring network. Cameras at key points on State Highways collect anonymised travel time data by tracking the time it takes for vehicles to move between two points — a technique known as journey time monitoring. This data is used to manage congestion, inform Variable Message Signs, and plan infrastructure improvements.
Toll Roads
New Zealand’s toll roads — including the Northern Gateway, Tauranga Eastern Link, and Takitimu Drive — rely entirely on ANPR for enforcement. There are no toll booths. Instead, gantry-mounted cameras capture the plate of every passing vehicle and cross-reference it with the NZ Transport Agency’s toll payment database. Vehicles with pre-paid accounts are passed through automatically; others receive an invoice by mail.
Local Councils & Parking Enforcement
Councils across New Zealand use ANPR to enforce on-street parking time limits. A vehicle-mounted camera drives slowly through streets, capturing and time-stamping every parked vehicle’s plate. If the same plate is still detected in the same zone after the permitted period, an enforcement notice can be issued — all without a parking warden needing to chalk tyres or return on foot.
Parking Services Limited — A New Zealand Example
A great New Zealand example of ANPR technology in action at the private enforcement level is Parking Services Limited (PSL) — a 100% Kiwi-owned and operated company that provides parking enforcement services to businesses, body corporates, commercial premises, and housing complexes nationwide.
PSL offers state-of-the-art ANPR cameras that can be fitted at the entry and exit points of a private car park. Every vehicle that enters or exits is photographed and its number plate recorded automatically. The system compares each plate against a whitelist of authorised vehicles — residents, employees, or pre-registered visitors. Any vehicle not on the whitelist can be automatically flagged, and an enforcement breach notice issued to the registered owner.
PSL describes the system as ideal for high-traffic zones or protecting designated parking spaces. Whitelists can be updated in real time via a touchpad system at reception or the building entrance, making it flexible for environments where authorised users change frequently — such as apartment buildings or medical centres.
For property owners who do not need a fully automated camera system, PSL also offers a simpler mobile-app-based enforcement service: a property manager photographs the offending vehicle and uploads the image, and PSL issues an enforceable $95 breach notice to the registered owner using NZTA register data. The company positions itself as the only nationwide New Zealand-owned parking enforcement company that retains its profits within New Zealand.
Border Control & Customs
At international ports of entry and domestic border checkpoints, ANPR cameras log every vehicle that crosses — providing an auditable record of vehicle movements and enabling rapid identification of vehicles flagged for inspection or associated with prohibited persons.
Global Applications
Internationally, ANPR is also used for congestion charging (most famously in London), airport security, stadium access, fuel station drive-off prevention, and even retail car park management by large supermarket chains. In the UK alone, the national ANPR network reads tens of millions of plates per day, with records stored for use in future criminal investigations.
Privacy and Ethical Considerations
The proliferation of ANPR has not been without controversy. Civil liberties groups have raised concerns about the mass collection of vehicle location data, arguing that at scale, ANPR creates a detailed record of citizens’ movements over time — a form of surveillance that, in many jurisdictions, requires no warrant or individual suspicion to initiate.
Key concerns include the retention of data on innocent vehicle movements, the potential for misuse by government or commercial operators, errors in plate reading leading to wrongful enforcement, and the lack of public awareness about just how comprehensively vehicle movements are being recorded.
In New Zealand, the use of ANPR data by government agencies is regulated under the Privacy Act 2020, with the NZTA motor vehicle register governed by specific rules around authorised access. Private operators like parking enforcement companies are also obligated to comply with privacy legislation when accessing NZTA register data to identify vehicle owners.
As with any powerful technology, the key lies in proportionate use, transparent policies, and meaningful public oversight.
Conclusion
ANPR cameras are a remarkable example of how a relatively niche technology — originally developed for British policing in the 1970s — has become embedded into the everyday infrastructure of modern life. From catching stolen cars on a motorway to managing who parks in a Wellington apartment building, ANPR quietly underpins a vast range of enforcement, access control, and traffic management functions.
For New Zealand businesses struggling with unauthorised car park use, services like Parking Services Limited bring this technology within reach at a practical and affordable level. For government agencies, ANPR provides a force multiplier — enabling a degree of road and vehicle oversight that would be impossible with human enforcement alone.
As AI and camera hardware continue to improve, ANPR systems will only become faster, more accurate, and more widely deployed. Understanding how they work — and the legitimate questions they raise about privacy — is increasingly important for anyone navigating the modern world.
Further Reading & References
• Wikipedia — Automatic Number Plate Recognition: en.wikipedia.org/wiki/Automatic_number-plate_recognition
• Parking Services Limited NZ (ANPR cameras): parkingservices.co.nz/number-plate-recognition-cameras
- Georges Car Media — ANPR video demonstration: tiktok.com/@georgescarmedia/video/7377324182226472225
- Can Private Companies Enforce Parking Fines in NZ?
