Automatic Number Plate Recognition (ANPR) technology (alternatively referred to as License Plate Recognition, depending on the region or country you are in) is an omnipresent force in today’s society. You may have seen ANPR signs at the roadside or at a car park’s entrance warning drivers that it is in use, but what is it exactly? In a nutshell, ANPR technology takes an image of a vehicle captured by a camera and, from that image, extracts details relating to its license plate using optical character recognition (OCR). This information can be processed or indexed in a database to create vehicle location data for cross referencing purposes. Some systems can also store a photograph of the driver.
In recent years, its use has become ingrained in many applications such as traffic monitoring studies, parking-lot access control and payment, police enforcement, and in some cities (such as my own) tracking cars that are in seemingly ever-expanding congestion charge zones. You could be forgiven for thinking ANPR was invented in this millennium but, in fact, its history stretches back to 1976, when it was developed by the UK’s Police Scientific Development Branch. To put that in perspective, that’s the same year Apple Computer Company was formed, the first commercially available laser printer was introduced by IBM, and the Sex Pistols topped the charts with their debut single “Anarchy in the UK.” However, early ANPR systems were expensive to deploy and maintain, and use was limited until the 1990s, when more cost-effective and easier-to-run solutions were pioneered.
How Does It Work?
Early ANPR systems were reliant on being connected to a local computer, on which video could be processed using OCR. These early systems were complex and required an ADSL line inside a local interface cabinet (plus main power) which would transmit data to a Back Office Facility (BOF). As processing became more widely available, cameras sent video to the local computer. Nowadays, the majority of systems have rolled all the image capture, image processing, and OCR within the camera enclosure proper, and some have internal SIMs and integrated routers sending data directly to the BOF. Data standards will depend on the BOF configuration, and can include diagnostic data of the camera function plus a heartbeat that allows the BOF to determine that the camera is functioning correctly.
As a vehicle approaches, the camera takes a series of images and stores them in a file. The camera measures the height of pixels in the vehicle registration mark (VRM) as it takes the images. The system will subsequently select the best frame in which the VRM is of a sufficient size for the OCR software, and that frame will be scanned and converted to a digital string and stored. Some ANPR systems take multiple images and retain a “favorite” frame that should return a highly accurate reading; others will only take one image at a certain position. The success of ANPR capture is wholly dependent on how reliable the system is and the correct positioning of the camera.
ANPR systems are not without their flaws. One recent mishap resulted in a British man being levied a fine for allegedly driving in the bus lane, but he was about 120 miles away at the time. As it transpired, the camera picked out the word “Knitter” displayed on the t-shirt of a pedestrian crossing the bus lane and the system read it as “KN19TER” (which happened to be the man’s car registration plate).
Where Is ANPR Used?
Advances in sensor engineering has led to some of the more advanced cameras being able to read multiple traffic lanes—up to 10 meters, simultaneously—with vehicles moving up to 140 miles per hour. Add to that the billions of plates being read a year and that is a whole lot of data being captured and stored. But for what purpose?
ANPR technology has found its way into many aspects of our daily lives, such as vehicle access control, delivery tracking, car park management, traffic and journey time analysis, tolling and road user billing, as well as traffic enforcement. ANPR cameras are often used at sites that require a higher level of security, with the technology working with pre-defined lists to deny or grant vehicles access depending on their number plate. Unsurprisingly, law enforcement agencies and the police are major users. Here are some of the main areas where ANPR is used:
- Law Enforcement: Since its inception, the use of ANPR is firmly rooted in modern-day policing. According to the National Police Chief’s Council, currently there are 13,000 ANPR cameras deployed UK-wide, submitting on average around 55 million ANPR “read” records to national ANPR systems daily. The NPCC states that ANPR technology is used to “detect, deter, and disrupt criminality.” This could mean anything from detecting uninsured and untaxed vehicles to locating stolen vehicles, tackling travelling criminals and terrorist activity, and providing crucial lines of enquiry and evidence in the investigation of crime.
- Traffic Monitoring: Agencies charged with operating, maintaining, and improving motorways and major roads can utilize ANPR information in a variety of ways. For example, data can be used to calculate average journey times and, in the case of smart motorways, to control the flow of traffic through variable speed limits. To determine journey times, an ANPR camera is set up to target a single lane of traffic (typically the middle lane on a three lane motorway). As a vehicle passes the camera, its number plate is read and converted into a non-unique reference number, known as a tag. The same number plate will generate the same tag every time it passes an ANPR camera, which enables journey times to be calculated when cross-referencing the location of each tag on the road network.
- Car Park Management: For parking companies, ANPR offers a very efficient mechanism for issuing parking tickets and makes easier payment processes, visibility of spaces, and ticketless systems possible. If set up correctly, timed photographs from cameras at the entrance and exit of a car park give an accurate recording of how long a vehicle has used the facilities. Whilst they are probably expensive to purchase, they are very efficient—operating 24 hours a day and identifying every single overnight stay in a car park.
- They are also proven to be very lucrative: research conducted in 2020 by Churchill Car Insurance found UK drivers paid parking fine fees adding up to nearly £3million a day, with the average penalty standing at £68.90.
What Is Next For ANPR?
The last five years have seen some major changes taking place. Taking car park management as an example, users would buy a camera that was pre-focused and pre-configured to take images over a 15-metre span. This would involve buying multiple cameras for use on differently sized sites. In recent years, the adoption of automatic varifocal lenses that can be controlled on the camera has meant that users can access and control a camera remotely and adjust the focus and zoom of it. This represented a dramatic change with a one-camera-fits-all solution. One significant, ongoing change is the image sensors themselves. A decade ago, cameras were limited to 1.8 meters of road width, but high-megapixel cameras now mean that 8 to 10 meters of road width can be observed with just one camera. Cameras can even recognize the number plates of cars travelling at speeds of 140 miles per hour if they’re compliant with the UK government’s standards.
Another interesting development is that cheaper alternatives to ANPR systems are becoming more prevalent, with several companies providing CCTV surveillance cameras and their own software. While ANPR cameras typically cost around £4,000 each, a decent high-end CCTV surveillance camera costs £1,500 plus another £800 for the ANPR software, and that can translate into big savings for small- to medium-sized companies.
ANPR technology has evolved and adapted with modern-day society, and has found new outlets and applications beyond policing and security. As networks of ANPR cameras continue to grow in size, we will see an interlinked web of smart cities up and down the country, with the technology providing critical, real-time information to transport groups and commuters. This will further help to reduce traffic delays, enforce low emission zones, and manage car parks on a city-wide scale.
Subscribers to our Office CompleteView service can log in to the InfoCenter to see more information on emerging “smart” technologies. Not a subscriber? No problem. Just send us an email at sales@keypointintelligence.com for more info.
