Automatic number-plate recognition
![](http://upload.wikimedia.org/wikipedia/commons/thumb/0/06/FrenchNumberPlates.jpg/250px-FrenchNumberPlates.jpg)
Automatic number plate recognition (ANPR) is a computer system that uses optical character recognition on images to read the licence plates on vehicles. Existing systems can scan number plates at around one per second on cars travelling up to 100 mph (160 kph). They can either use existing closed-circuit television or road-rule enforcement cameras, or ones specifically designed for the task.
ANPR can be used to store the images captured by the cameras as well as the text from the licence plate, with some configurable to store a photograph of the driver. Systems commonly use infrared lighting to allow the camera to take the picture at any time of day. They also tend to be country-specific due to the variation of plates internationally.
The software aspect of the system runs on a standard computer running Microsoft Windows or Linux and can be linked to existing applications or databases.
Other names
ANPR is sometimes known by various other terms:
- Automatic vehicle identification (AVI)
- Car plate recognition (CPR)
- Licence plate recognition (LPR)
Technology
![](http://upload.wikimedia.org/wikipedia/commons/3/3c/Dutch_license_plate_segment.jpg)
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 licence plate arrangements use variations in font sizes and positioning – ANPR systems must be able to cope with such differences in order to be truly effective. More complicated systems can cope with international variants, though most existing programs are tailored to each country individually.
The cameras used can include 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.
Algorithms
![](http://upload.wikimedia.org/wikipedia/commons/thumb/9/9c/California_license_plate_ANPR.png/250px-California_license_plate_ANPR.png)
There are five primary algorithms that the software requires for identifying a licence plate:
- Plate localisation – 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
- Normalisation – adjusts the brightness and contrast of the image
- Character segmentation – finds the individual characters on the plates
- Optical character recognition
The complexity of each of these subsections of the program determines the accuracy of the system. During the third phase (normalisation) 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 noise on the image.
Difficulties
There are a number of possible difficulties that the software must be able to cope with. These include:
- Poor image resolution, usually because the plate is too far away but sometimes because of a poor quality image
- Blurry images, particularly motion blur and most likely on mobile units
- Poor lighting and low contrast due to overexposure, reflection or shadows
- An object obscuring (part of) the plate, quite often a tow bar, or a dirty plate
- A different font, popular for vanity plates though illegal in many countries
- Driver circumvention techniques
While some of these problems can be corrected within the software it is primarily down to the hardware side of the system to circumvent these difficulties. If a camera is placed high up it might help with objects obscuring the plate, including other vehicles if they are driving close together, but introduces and increases other problems such as the skew of the plate.
Many countries now use licence plates that are retroreflective. This returns the light back to the source and thus improves the contrast of the image. A camera that makes use of infrared imaging (with a normal colour filter over the lens and an infrared light-source next to it) benefits greatly from this as the infrared waves are reflected back from the plate. The characters on the plate are not reflective, therefore giving a high level of contrast in any lighting conditions. 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 camera working together.
![](http://upload.wikimedia.org/wikipedia/commons/thumb/a/a2/California_license_plate_blur.png/250px-California_license_plate_blur.png)
To avoid blurring it is ideal to have the shutter speed of a dedicated camera set to 1/1000th of a second. Anything more than this, especially when the camera is quite high up relative to the vehicle, could result in an image which is too blurred to read using the OCR software. 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 can cope with traffic moving up to 40 mph (64 kph) and 1/250 up to 5 mph (8 kph).
Some small-scale systems allow for some errors in the licence plate. When used for giving specific vehicles access to an barriered area the decision may be made to have an acceptable error rate of one character. This is because the likelihood of an unauthorised car having such a similar licence plate is seen as quite small. However, this level of inaccuracy would not be acceptable in most applications of an ANPR system.
Some companies have sold a special transparent spray-on coating claiming that it "tricks" the cameras. The spray 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.
Police enforcement
![](http://upload.wikimedia.org/wikipedia/commons/thumb/8/80/Closed.circuit.twocameras.arp.750pix.jpg/250px-Closed.circuit.twocameras.arp.750pix.jpg)
After the licence plate has been identified it can then be cross-referenced against a police database. The primary objectives of this are to identify vehicles that have been stolen, used in a crime or are in violation of some other law. Some systems are also linked to insurance databases to monitor if the vehicle is currently insured.
Project Laser in the United Kingdom
In 2002 the Vehicle and Operator Services Agency (VOSA) was given funding by the British Home Office to work with the Police Standards Unit and develop "Project Laser". With the aim of running the ANPR system nationwide, it was initially trialled by nine police forces and ran between 30 September 2002 and March 2003. Those police forces were:
- Greater Manchester
- North Wales
- Avon and Somerset
- Northamptonshire
- The Metropolitan Police Service
- Kent
- West Yorkshire
- Staffordshire
- West Midlands.
The second phase of the project ran between 1 June 2003 and 31 June 2004 and involved 23 police forces in total. The DVLA is also involved with Project Laser, using the system to gather details on unregistered and unlicenced vehicles and those without a valid MOT certificate or insurance cover.
- "Eventually the database will link to most CCTV systems in town centres, meaning that all vehicles filmed on one of the many cameras protecting Bedford High Street, for instance, can be checked against the database and the movements of wanted cars traced to help with serious crime investigations."
- – Bedfordshire Police [1]
The project was seen as a success despite a Home Office report showing that the Driver and Vehicle Licensing Agency (DVLA) trial had an error rate of up to 40%, with claims that the system was contributing "in excess of 100 arrests per officer per year – ten times the national average." [2] Further findings went on to show that the error rate dropped to 5% when infrared systems and updated software were used.
During the second phase of the project around 28 million number plates were spotted in total, with 1.1 million (3.9%) of these matching an entry in one of the databases. 180,543 vehicles were stopped (101,775 directly because of the ANPR system), leading to 13,499 arrests (7.5% of the total) and the issue of 50,910 fines (28.2%). 1,152 stolen vehicles (worth £7.5 million in total), £380,000 worth of drugs and £640,000 worth of stolen goods were also recovered. The primary goal of the second phase was, however, to see how well the costs of the ANPR system could be covered. The final conclusion was that less than 10% of the expenditure incurred was recouped, with the Home Office claiming that the failure of drivers to pay fines contributed to this low figure, and continued to recommend the system be deployed throughout the UK.
Traffic charge zones
![](http://upload.wikimedia.org/wikipedia/commons/thumb/e/ec/London-cc-mobile-l.jpg/250px-London-cc-mobile-l.jpg)
The London Congestion Charge is an example of a system that charges motorists entering a payment area. Transport for London (TfL) uses ANPR systems and charges motorists a daily fee of £5 if they enter, leave or move around within the London Congestion Charge zone between 7 a.m. and 6:30 p.m., Monday to Friday.
Two hundred and thirty CCTV-style cameras, of which 180 are installed at the edge of the zone, are currently in use. In addition to the 180 cameras on the edge of the zone, there are fifty further cameras placed within it. These cameras are intended to pick up cars that are missed on entry and/or exit and those that are moving solely within the zone. There are also a number of mobile camera units which may be deployed anywhere in the zone.
It is estimated that around 98% of vehicles moving within the zone are caught on camera. The video streams are transmitted to a data centre located in Central London where the ANPR software deduces the registration plate of the vehicle. A second data centre provides a backup location for image data.
Both front and back number plates are being captured, on vehicles going both in and out – this gives up to four chances to capture the number plates of a vehicle entering and exiting the zone. This list is then compared with a list of cars whose owners/operators have paid to enter the zone – those that have not paid are fined. The registered owner of such a vehicle is looked up in a database provided by the DVLA.
ANPR alongside alternative systems
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 that opt to rent a transponder for C$2.00 per month are not charged the "Video Toll Charge" of C$3.45 for using the road, with heavy vehicles (those with a gross weight of over 5,000 kilograms) being required to use one. Using either system, users of the highway are notified of the usage charges by post.
Controversy
The introduction of ANPR systems has led to fears of misidentification and the furthering of Big Brother information. Systems are still fallible with one critic of the London Congestion Charge scheme noting "Misread plate after misread plate appeared on the screen – of every 10 that appeared at least four were incorrect." [3] This can 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 offence.
Other concerns include the storage of information contravening the Data Protection Act 1984 along with the freedom of information and similar legislation. The laws in the UK are strict if a system uses CCTV footage and can identify individuals.
Other uses
ANPR systems may also be used by:
- Toll booths and border crossings
- Filling stations to log when a driver drives away without paying
- Car parks or road entry systems to control access
- A marketing tool to log patterns of use
- Traffic management systems, which determine traffic flow using the time it takes vehicles to pass two ANPR sites.
References
- ANPR by the Police Information Technology Organisation (PITO), referenced 28 March 2005
- ANPR at the Police Standards Unit section of PoliceReform.gov.uk, referenced 28 March 2005
- Lettice, John. 24 March 2005. "No hiding place? UK number plate cameras go national" at The Register, referenced 28 March 2005
- "Operation Mermaid – National ANPR Day" by Bedfordshire Police, referenced 28 March 2005
- Hofman, Yoram. "License Plate Recognition - A Tutorial", referenced 28 March 2005
- DVLA business plan, referenced 28 March 2005
- Millar, Chris. 20 February 2003. "Exposed: Ken's camera spies" at ThisIsLondon.com, referenced 28 March, 2005
- Lettice, John. 21 February 2003. "The London charge zone, the DP Act, and MS .NET" at The Register, referenced 28 March, 2005
- "Driving crime down", a PDF file from the Home Office, referenced 29 March 2005
- Constant, Mike. CCTV Information – ANPR, referenced 30 March 2005
- 407 ETR page about transponders, referenced 31 March 2005
External links
Companies and agencies using and providing ANPR systems:
- Vehicle and Operator Services Agency (VOSA)
- Police Information Technology Organisation (PITO)
- 407 ETR
- Transport for London
- Trafficmaster
News and reports:
- "Number plate recognition poised for national UK rollout" at The Register
- "Number plate scan to be extended" at BBC News
- "No Smiles for the Camera" at SourceUK.net
- "London Congestion Charge CCTV privacy concerns" at spy.org.uk
- "Car Cloning" at BBC Inside Out
Research:
- Plate Recognition at PhotoCop.com
- "A Real-time vehicle License Plate Recognition (LPR)" at visl.technion.ac.il
- "An Approach To Licence Plate Recognition" – a PDF file describing a University of Calgary project that looks at plate location in raster images
Information from developers of ANPR systems: