automatic number plate recognition

Other names

ANPR is also known by various other terms:

Automatic (or automated) license-plate recognition (ALPR)

Automatic (or automated) license-plate reader (ALPR)

Automatic vehicle identification (AVI)

Danish: Automatisk nummerpladegenkendelse, lit. 'Automatic number plate recognition' (ANPG)

Car-plate recognition (CPR)

License-plate recognition (LPR)

French: Lecture automatique de plaques d'immatriculation, lit. 'Automatic reading of registration plates' (LAPI)

Mobile license-plate reader (MLPR)

Vehicle license-plate recognition (VLPR)

Vehicle recognition identification (VRI)

Development

ANPR was invented in 1976 at the Police Scientific Development Branch in Britain. Prototype systems were working by 1979, and contracts were awarded to produce industrial systems, first at EMI Electronics, and then at Computer Recognition Systems (CRS, now part of Jenoptik) in Wokingham, UK. Early trial systems were deployed on the A1 road and at the Dartford Tunnel. The first arrest through detection of a stolen car was made in 1981. However, ANPR did not become widely used until new developments in cheaper and easier-to-use software were pioneered during the 1990s. The collection of ANPR data for future use (i.e., in solving then-unidentified crimes) was documented in the early 2000s. The first documented case of ANPR being used to help solve a murder occurred in November 2005, in Bradford, UK, where ANPR played a vital role in locating and subsequently convicting the killers of Sharon Beshenivsky.

Components

The software aspect of the system runs on standard home computer hardware and can be linked to other applications or databases. It first uses a series of image manipulation techniques to detect, normalize and enhance the image of the number plate, and then optical character recognition (OCR) to extract the alphanumerics of the license plate. ANPR systems are generally deployed in one of two basic approaches: one allows for the entire process to be performed at the lane location in real-time, and the other transmits all the images from many lanes to a remote computer location and performs the OCR process there at some later point in time. When done at the lane site, the information captured of the plate alphanumeric, date-time, lane identification, and any other information required is completed in approximately 250 milliseconds. This information can easily be transmitted to a remote computer for further processing if necessary, or stored at the lane for later retrieval. In the other arrangement, there are typically large numbers of PCs used in a server farm to handle high workloads, such as those found in the London congestion charge project. Often in such systems, there is a requirement to forward images to the remote server, and this can require larger bandwidth transmission media.

Technology

ANPR uses optical character recognition (OCR) on images taken by cameras. When Dutch vehicle registration plates switched to a different style in 2002, one of the changes made was to the font, introducing small gaps in some letters (such as P and R) to make them more distinct and therefore more legible to such systems. Some license plate arrangements use variations in font sizes and positioning – ANPR systems must be able to cope with such differences to be truly effective. More complicated systems can cope with international variants, though many programs are individually tailored to each country.

The cameras used can be existing road-rule enforcement or closed-circuit television cameras, as well as mobile units, which are usually attached to vehicles. Some systems use infrared cameras to take a clearer image of the plates.

During the 1990s, significant advances in technology took automatic number-plate recognition (ANPR) systems from limited expensive, hard-to-set-up, fixed-based applications to simple "point and shoot" mobile ones. This was made possible by the creation of software that ran on cheaper PC-based, non-specialist hardware that also no longer needed to be given the pre-defined angles, direction, size and speed in which the plates would be passing the camera's field of view. Further scaled-down components at lower price points led to a record number of deployments by law enforcement agencies globally. Smaller cameras with the ability to read license plates at higher speeds, along with smaller, more durable processors that fit in the trunks of police vehicles, allowed law enforcement officers to patrol daily with the benefit of license plate reading in real time, when they can interdict immediately.

Despite their effectiveness, there are noteworthy challenges related with mobile ANPRs. One of the biggest is that the processor and the cameras must work fast enough to accommodate relative speeds of more than 160 km/h (100 mph), a likely scenario in the case of oncoming traffic. This equipment must also be very efficient since the power source is the vehicle electrical system, and equipment must have minimal space requirements.

Relative speed is only one issue that affects the camera's ability to read a license plate. Algorithms must be able to compensate for all the variables that can affect the ANPR's ability to produce an accurate read, such as time of day, weather and angles between the cameras and the license plates. A system's illumination wavelengths can also have a direct impact on the resolution and accuracy of a read in these conditions.

Algorithms

There are seven primary algorithms that the software requires for identifying a license plate:

Plate localization – responsible for finding and isolating the plate on the picture

Plate orientation and sizing – compensates for the skew of the plate and adjusts the dimensions to the required size

Normalization – adjusts the brightness and contrast of the image

Character segmentation – finds the individual characters on the plates

Optical character recognition

Syntactical/Geometrical analysis – check characters and positions against country-specific rules

The averaging of the recognised value over multiple fields/images to produce a more reliable or confident result, especially given that any single image may contain a reflected light flare, be partially obscured, or possess other obfuscating effects.

The complexity of each of these subsections of the program determines the accuracy of the system. During the third phase (normalization), some systems use edge detection techniques to increase the picture difference between the letters and the plate backing. A median filter may also be used to reduce the visual noise on the image.

Contemporary ANPR systems use multiple data sources and analytical techniques that go beyond simple number plate recognition. Weigh-in-Motion uses ANPR cameras and AI analytical techniques to calculate the weight of a vehicle and to alert if the weight is too high for the vehicle and conditions.

There are a number of possible difficulties that the software must be able to cope with. These include:

Imaging hardware

At the front end of any ANPR system is the imaging hardware which captures the image of the license plates. The initial image capture forms a critically important part of the ANPR system which, in accordance to the garbage in, garbage out principle of computing, will often determine the overall performance.

License plate capture is typically performed by specialized cameras designed specifically for the task, although new software techniques are being implemented that support any IP-based surveillance camera and increase the utility of ANPR for perimeter security applications. Factors which pose difficulty for license plate imaging cameras include the speed of the vehicles being recorded, varying level of ambient light, headlight glare and harsh environmental conditions. Most dedicated license plate capture cameras will incorporate infrared illumination in order to solve the problems of lighting and plate reflectivity.

Many countries now use license plates that are retroreflective. This returns the light back to the source and thus improves the contrast of the image. In some countries, the characters on the plate are not reflective, giving a high level of contrast with the reflective background in any lighting conditions. A camera that makes use of active infrared imaging (with a normal colour filter over the lens and an infrared illuminator next to it) benefits greatly from this as the infrared waves are reflected back from the plate. This is only possible on dedicated ANPR cameras, however, and so cameras used for other purposes must rely more heavily on the software capabilities. Further, when a full-colour image is required as well as use of the ANPR-retrieved details, it is necessary to have one infrared-enabled camera and one normal (colour) camera working together.

To avoid blurring it is ideal to have the shutter speed of a dedicated camera set to 1⁄1000 of a second. It is also important that the camera use a global shutter, as opposed to rolling shutter, to assure that the taken images are distortion-free. Because the car is moving, slower shutter speeds could result in an image which is too blurred to read using the OCR software, especially if the camera is much higher up than the vehicle. In slow-moving traffic, or when the camera is at a lower level and the vehicle is at an angle approaching the camera, the shutter speed does not need to be so fast. Shutter speeds of 1⁄500 of a second can cope with traffic moving up to 65 km/h (40 mph) and 1⁄250 of a second up to 8 km/h (5 mph). License plate capture cameras can produce usable images from vehicles traveling at 190 km/h (120 mph).

Usage

Law enforcement

Several State Police Forces, and the Department of Justice (Victoria) use both fixed and mobile ANPR systems. The New South Wales Police Force Highway Patrol were the first to trial and use a fixed ANPR camera system in Australia in 2005. In 2009 they began a roll-out of a mobile ANPR system (known officially as MANPR) with three infrared cameras fitted to its Highway Patrol fleet. The system identifies unregistered and stolen vehicles as well as disqualified or suspended drivers as well as other 'persons of interest' such as persons having outstanding warrants.

The city of Mechelen uses an ANPR system since September 2011 to scan all cars crossing the city limits (inbound and outbound). Cars listed on 'black lists' (no insurance, stolen, etc.) generate an alarm in the dispatching room, so they can be intercepted by a patrol.

As of early 2012, 1 million cars per week are automatically checked in this way.

Federal, provincial, and municipal police services across Canada use automatic licence plate recognition software; they are also used on certain toll routes and by parking enforcement agencies. Laws governing usage of information thus obtained use of such devices are mandated through various provincial privacy acts.

The technique is tested by the Danish police. It has been in permanent use since mid 2016.

180 gantries over major roads have been built throughout the country. These together with a further 250 fixed cameras is to enable a levy of an eco tax on lorries over 3.5 tonnes. The system is currently being opposed and whilst they may be collecting data on vehicles passing the cameras, no eco tax is being charged.

On 11 March 2008, the Federal Constitutional Court of Germany ruled that some areas of the laws permitting the use of automated number plate recognition systems in Germany violated the right to privacy. More specifically, the court found that the retention of any sort of information (i.e., number plate data) which was not for any pre-destined use (e.g., for use tracking suspected terrorists or for enforcement of speeding laws) was in violation of German law.

Average-speed cameras

ANPR is used for speed limit enforcement in Australia, Austria, Belgium, Dubai (UAE), France, Ireland, Italy, The Netherlands, Spain, South Africa, the UK, and Kuwait.

This works by tracking vehicles' travel time between two fixed points, and calculating the average speed. These cameras are claimed to have an advantage over traditional speed cameras in maintaining steady legal speeds over extended distances, rather than encouraging heavy braking on approach to specific camera locations and subsequent acceleration back to illegal speeds.

In Italian highways there is a monitoring system named Tutor covering more than 2,500 km (1,600 miles) (2012). The Tutor system is also able to intercept cars while changing lanes. The Tutor or Safety Tutor is a joint project between the motorway management company, Autostrade per l'Italia, and the State Police. Over time it has been replaced by other versions for example the SICVe-PM where PM stands for PlateMatching and by the SICVe Vergilius. In addition to this average speed monitoring system, there are others Celeritas and T-Expeed v.2.

Average speed cameras (trajectcontrole) are in place in the Netherlands since 2002. As of July 2009, 12 cameras were operational, mostly in the west of the country and along the A12. Some of these are divided in several "sections" to allow for cars leaving and entering the motorway.

A first experimental system was tested on a short stretch of the A2 in 1997 and was deemed a big success by the police, reducing overspeeding to 0.66%, compared to 5 to 6% when regular speed cameras were used at the same location. The first permanent average speed cameras were installed on the A13 in 2002, shortly after the speed limit was reduced to 80 km/h (50 mph) to limit noise and air pollution in the area. In 2007, average speed cameras resulted in 1.7 million fines for overspeeding out of a total of 9.7 million. According to the Dutch Attorney General, the average number of violation of the speed limits on motorway sections equipped with average speed cameras is between 1 and 2%, compared to 10 to 15% elsewhere.

One of the most notable stretches of average speed cameras in the UK is found on the A77 road in Scotland, with 32 miles (51 km) being monitored between Kilmarnock and Girvan. In 2006 it was confirmed that speeding tickets could potentially be avoided from the 'SPECS' cameras by changing lanes and the RAC Foundation feared that people may play "Russian Roulette" changing from one lane to another to lessen their odds of being caught; however, in 2007 the system was upgraded for multi-lane use and in 2008 the manufacturer described the "myth" as "categorically untrue". There exists evidence that implementation of systems such as SPECS has a considerable effect on the volume of drivers travelling at excessive speeds; on the stretch of road mentioned above (A77 Between Glasgow and Ayr) there has been noted a "huge drop" in speeding violations since the introduction of a SPECS system.

Crime deterrent

Recent innovations have contributed to the adoption of ANPR for perimeter security and access control applications at government facilities. Within the US, "homeland security" efforts to protect against alleged "acts of terrorism" have resulted in adoption of ANPR for sensitive facilities such as embassies, schools, airports, maritime ports, military and federal buildings, law enforcement and government facilities, and transportation centers. ANPR is marketed as able to be implemented through networks of IP based surveillance cameras that perform "double duty" alongside facial recognition, object tracking, and recording systems for the purpose of monitoring suspicious or anomalous behavior, improving access control, and matching against watch lists. ANPR systems are most commonly installed at points of significant sensitivity, ingress or egress. Major US agencies such as the Department of Homeland Security, the Department of Justice, the Department of Transportation and the Department of Defense have purchased ANPR for perimeter security applications. Large networks of ANPR systems are being installed by cities such as Boston, London and New York City to provide citywide protection against acts of terrorism, and to provide support for public gatherings and public spaces.

The Center For Evidence-Based Crime Policy in George Mason University identifies the following randomized controlled trials of automatic number-plate recognition technology as very rigorous.

Enterprise security and services

In addition to government facilities, many private sector industries with facility security concerns are beginning to implement ANPR solutions. Examples include casinos, hospitals, museums, parking facilities, and resorts. In the US, private facilities typically cannot access government or police watch lists, but may develop and match against their own databases for customers, VIPs, critical personnel or "banned person" lists. In addition to providing perimeter security, private ANPR has service applications for valet / recognized customer and VIP recognition, logistics and key personnel tracking, sales and advertising, parking management, and logistics (vendor and support vehicle tracking).

Traffic control

Many cities and districts have developed traffic control systems to help monitor the movement and flow of vehicles around the road network. This had typically involved looking at historical data, estimates, observations and statistics, such as:

Car park usage

Pedestrian crossing usage

Number of vehicles along a road

Areas of low and high congestion

Frequency, location and cause of road works

CCTV cameras can be used to help traffic control centres by giving them live data, allowing for traffic management decisions to be made in real-time. By using ANPR on this footage it is possible to monitor the travel of individual vehicles, automatically providing information about the speed and flow of various routes. These details can highlight problem areas as and when they occur and help the centre to make informed incident management decisions.

Some counties of the United Kingdom have worked with Siemens Traffic to develop traffic monitoring systems for their own control centres and for the public. Projects such as Hampshire County Council's ROMANSE provide an interactive and real-time website showing details about traffic in the city. The site shows information about car parks, ongoing road works, special events and footage taken from CCTV cameras. ANPR systems can be used to provide average point-to-point journey times along particular routes, which can be displayed on a variable-message sign (VMS) giving drivers the ability to plan their route. ROMANSE also allows travellers to see the current situation using a mobile device with an Internet connection (such as WAP, GPRS or 3G), allowing them to view mobile device CCTV images within the Hampshire road network.

The UK company Trafficmaster has used ANPR since 1998 to estimate average traffic speeds on non-motorway roads without the results being skewed by local fluctuations caused by traffic lights and similar. The company now operates a network of over 4000 ANPR cameras, but claims that only the four most central digits are identified, and no numberplate data is retained.

Electronic toll collection

Ontario's 407 ETR highway uses a combination of ANPR and radio transponders to toll vehicles entering and exiting the road. Radio antennas are located at each junction and detect the transponders, logging the unique identity of each vehicle in much the same way as the ANPR system does. Without ANPR as a second system it would not be possible to monitor all the traffic. Drivers who opt to rent a transponder for CA$2.55 (US$1.92) per month are not charged the "Video Toll Charge" of CA$3.6 (US$2.71) for using the road, with heavy vehicles (those with a gross weight of over 5,000 kg or 5.5 short tons) being required to use one. Using either system, users of the highway are notified of the usage charges by post.

There are numerous other electronic toll collection networks which use this combination of Radio frequency identification and ANPR. These include:

The Golden Gate Bridge in San Francisco, California, which began using an all-electronic tolling system combining Fastrak and ANPR on 27 March 2013

NC Quick Pass for the Interstate 540 (North Carolina) Triangle Expressway in Wake County, North Carolina

Bridge Pass for the Saint John Harbour Bridge in Saint John, New Brunswick

Quickpass at the Golden Ears Bridge, crossing the Fraser River between Langley and Maple Ridge

e-TAG, all Australian toll roads

FasTrak in California, United States

Highway 6 in Israel

Tunnels in Hong Kong

Autopista Central in Santiago, Chile

E-ZPass in New York, New Jersey, Pennsylvania, Massachusetts (as Fast Lane until 2012), Virginia (formerly Smart Tag), and other states. Maryland Route 200 uses a combination of E-ZPass and ANPR.

Private use

Several UK companies and agencies use ANPR systems. These include Vehicle and Operator Services Agency (VOSA), Driver and Vehicle Licensing Agency (DVLA) and Transport for London.

Other uses

ANPR systems may also be used for/by:

Section control, to measure average vehicle speed over longer distances

Border crossings

Automobile repossessions

Petrol stations to log when a motorist drives away without paying for their fuel

A marketing tool to log patterns of use

Targeted advertising, a-la "Minority Report"-style billboards

Traffic management systems, which determine traffic flow using the time it takes vehicles to pass two ANPR sites

Analyses of travel behaviour (route choice, origin-destination etc.) for transport planning purposes

Drive-through customer recognition, to automatically recognize customers based on their license plate and offer them the items they ordered the last time they used the service

To assist visitor management systems in recognizing guest vehicles

Police and auxiliary police

Car parking companies

To raise or lower automatic bollards

Hotels

Enforcing Move over laws for emergency vehicles

Automated emissions testing

Challenges

Circumvention

Vehicle owners have used a variety of techniques in an attempt to evade ANPR systems and road-rule enforcement cameras in general. One method increases the reflective properties of the lettering and makes it more likely that the system will be unable to locate the plate or produce a high enough level of contrast to be able to read it. This is typically done by using a plate cover or a spray, though claims regarding the effectiveness of the latter are disputed. In most jurisdictions, the covers are illegal and covered under existing laws, while in most countries there is no law to disallow the use of the sprays. Other users have attempted to smear their license plate with dirt or utilize covers to mask the plate.

Novelty frames around Texas license plates were made illegal in Texas on 1 September 2003 by Texas Senate Bill 439 because they caused problems with ANPR devices. That law made it a Class C misdemeanor (punishable by a fine of up to US$200), or Class B (punishable by a fine of up to US$2,000 and 180 days in jail) if it can be proven that the owner did it to deliberately obscure their plates. The law was later clarified in 2007 to allow novelty frames.

If an ANPR system cannot read the plate, it can flag the image for attention, with the human operators looking to see if they are able to identify the alphanumerics.

In 2013 researchers at Sunflex Zone Ltd created a privacy license plate frame that uses near infrared light to make the license plate unreadable to license plate recognition systems.

Privacy concerns and misuse

The introduction of ANPR systems has led to fears of misidentification and the furthering of 1984-style surveillance. In the United States, some such as Gregg Easterbrook oppose what they call "machines that issue speeding tickets and red-light tickets" as the beginning of a slippery slope towards an automated justice system:

"A machine classifies a person as an offender, and you can't confront your accuser because there is no accuser... can it be wise to establish a principle that when a machine says you did something illegal, you are presumed guilty?"

Similar criticisms have been raised in other countries. Easterbrook also argues that this technology is employed to maximize revenue for the state, rather than to promote safety.

The electronic surveillance system produces tickets which in the US are often in excess of $100, and are virtually impossible for a citizen to contest in court without the help of an attorney. The revenues generated by these machines are shared generously with the private corporation that builds and operates them, creating a strong incentive to tweak the system to generate as many tickets as possible.

Older systems had been notably unreliable; in the UK this has been known to lead to charges being made incorrectly with the vehicle owner having to pay £10 in order to be issued with proof (or not) of the offense. Improvements in technology have drastically decreased error rates, but false accusations are still frequent enough to be a problem.

Perhaps the best known incident involving the abuse of an ANPR database in North America is the case of Edmonton Sun reporter Kerry Diotte in 2004. Diotte wrote an article critical of Edmonton police use of traffic cameras for revenue enhancement, and in retaliation was added to an ANPR database of "high-risk drivers" in an attempt to monitor his habits and create an opportunity to arrest him. The police chief and several officers were fired as a result, and The Office of the Privacy Commissioner of Canada expressed public concern over the "growing police use of technology to spy on motorists."

Plate inconsistency and jurisdictional differences

Many ANPR systems claim accuracy when trained to match plates from a single jurisdiction or region, but can fail when trying to recognize plates from other jurisdictions due to variations in format, font, color, layout, and other plate features. Some jurisdictions (particularly in the US) offer vanity or affinity plates, which can create many variations within a single jurisdiction.

From time to time, US states will make significant changes in their license plate protocol that will affect OCR accuracy. They may add a character or add a new license plate design. ALPR systems must adapt to these changes quickly in order to be effective. Another challenge with ALPR systems is that some states have the same license plate protocol. For example, more than one state uses the standard three letters followed by four numbers. So each time the ALPR systems alarms, it is the user's responsibility to make sure that the plate which caused the alarm matches the state associated with the license plate listed on the in-car computer. For maximum effectiveness, an ANPR system should be able to recognize plates from any jurisdiction, and the jurisdiction to which they are associated, but these many variables make such tasks difficult.

Currently at least one US ANPR provider (PlateSmart) claims their system has been independently reviewed as able to accurately recognize the US state jurisdiction of license plates, and one European ANPR provider claims their system can differentiate all EU plate jurisdictions.

Accuracy and measurement of ANPR system performance

A few ANPR software vendors publish accuracy results based on image benchmarks. These results may vary depending on which images the vendor has chosen to include in their test. In 2017, Sighthound reported a 93.6% accuracy on a private image benchmark. In 2017, OpenALPR reported accuracy rates for their commercial software in the range of 95-98% on a public image benchmark. April 2018 research from Brazil's Federal University of Paraná and Federal University of Minas Gerais obtained a recognition rate of 93.0% for OpenALPR and 89.8% for Sighthound, running both on the SSIG dataset; and a rate of 93.5% for a system of their own design based on the YOLO object detector, also using the SSIG dataset. Testing a "more realistic scenario" involving both plate and reader moving, the researchers obtained rates of less than 70% for the two commercial systems and 78.3% for their own.

Limitations of legacy LPR systems

In some contexts, the term legacy LPR is used to describe older licence plate recognition systems that rely solely on black-and-white optical character recognition (OCR) of the plate itself, without capturing broader contextual data such as vehicle type, colour, or surrounding road conditions. These systems can be prone to higher false-positive rates, may struggle with new plate formats, and are generally limited in their ability to detect complex violations such as wrong-way parking, misuse of loading zones, or the presence (or absence) of parking permits visible only inside a vehicle's windscreen.