Automatic number number plate recognition recognition From Wikipedia, the free encyclopedia
The system must be able to deal with different styles of license of license plates
License-plate recognition process
names below) is a mass Automatic number plate recognition (ANPR; see also other names below) sureillance method that uses optical character recognition on images to read ehicle registration sureillance method plates!! They can use e"isting closed-circuit teleision or road-rule enforcement cameras, plates cameras, or ones specifically designed for the task! They are used by b y arious police forces and as a method
of electronic electronic toll collection on collection on pay-per-use roads and roads and cataloging the moements of traffic or indiiduals! #$%& can be used to store store the images captured by the the cameras as well as the te"t te"t from the license plate, with some configurable to store a photograph of the drier! 'ystems commonly *+* +** * use infrared infraredllighting to allow the camera to take the picture at any time of the day! day!* #$%& technology tends to be region-specific, owing to plate pl ate ariation from place to place! oncerns about these systems hae centered on priacy fears of goernment tracking citi.ens/ moements, misidentification, high error rates, and increased goernment spending! Contents
hide hide** •
0tym 0tymolo ology gy
•
+ 1eelo 1eelopment pment history
•
omp ompone onents nts
•
2 Techn Technol olog ogyy o
2! #$%& in mobile systems
•
3 #lg #lgori orithms thms
•
4 1iff 1ifficu iculti lties es
•
5 6mag 6maging ing hard hardware ware
•
7 ircum ircumention ention techni8ues
•
9 %olice enforcement enforcemen t o
9! #us #ustral tralia ia
o
9!+ :el :elgiu gium m
o
9! 1e 1enm nmark ark
o
9!2 Fra France nce
o
9!3 e erma rmany ny
o
9!4
o
9!5 Tur Turke keyy
o
9!7 =kra =kraine ine
of electronic electronic toll collection on collection on pay-per-use roads and roads and cataloging the moements of traffic or indiiduals! #$%& can be used to store store the images captured by the the cameras as well as the te"t te"t from the license plate, with some configurable to store a photograph of the drier! 'ystems commonly *+* +** * use infrared infraredllighting to allow the camera to take the picture at any time of the day! day!* #$%& technology tends to be region-specific, owing to plate pl ate ariation from place to place! oncerns about these systems hae centered on priacy fears of goernment tracking citi.ens/ moements, misidentification, high error rates, and increased goernment spending! Contents
hide hide** •
0tym 0tymolo ology gy
•
+ 1eelo 1eelopment pment history
•
omp ompone onents nts
•
2 Techn Technol olog ogyy o
2! #$%& in mobile systems
•
3 #lg #lgori orithms thms
•
4 1iff 1ifficu iculti lties es
•
5 6mag 6maging ing hard hardware ware
•
7 ircum ircumention ention techni8ues
•
9 %olice enforcement enforcemen t o
9! #us #ustral tralia ia
o
9!+ :el :elgiu gium m
o
9! 1e 1enm nmark ark
o
9!2 Fra France nce
o
9!3 e erma rmany ny
o
9!4
o
9!5 Tur Turke keyy
o
9!7 =kra =kraine ine
•
o
9!9 =nited >ingdom
o
9!? =ni =nited ted 'tat 'tates es
o
9! 'au 'audi di #rab #rabia ia
o
9!+ 'we 'weden den
? #erage-s #erage-speed peed cameras o
?! ? ! 6ta 6taly ly
o
?!+ The $etherl $etherlands ands
o
?! ? ! =>
•
ri rime me dete deterren rrentt
•
+ 0nterprise security and seric serices es
•
Traffi Trafficc cont control rol
•
2 0lectronic toll collec collection tion o
2! Toll road roadss
o
2!+ %ort %ortuga ugall
o
2! harge .ones @ the London congestion conges tion charge
o
2!2 'we 'weden den
•
3 =s =sag age e
•
4 ha halle llenge ngess o
4! on ontro troers ersyy
o
4!+ %late incons inconsistenc istencyy and Aurisdictional Aurisdiction al differenc differences es
o
4! #ccuracy and measurement measureme nt of #$%& system performance
•
5 Bther uses
•
7 &elated research society
•
9 'ee al also so
•
+? &eferences
Etymology edit* #$%& is sometimes known by arious other termsC •
Automatic license-plate recognition (#L%&)
•
Automatic license-plate reader (#L%&)
•
Automatic vehicle identification (#D6)
•
Car plate recognition (%&)
•
License-plate recognition (L%&)
•
Lecture automatique de plaques d'immatriculation (L#%6)
•
Mobile license-plate reader (EL%&)
Development history edit* #$%& was inented in 954 at the %olice 'cientific 1eelopment :ranch in the =>!citation needed * %rototype systems were working by 959, and contracts were let to produce industrial systems, first at 0E6 0lectronics, and then at omputer &ecognition 'ystems (&') in Wokingham, =>! 0arly trial systems were deployed on the # road and at the 1artford Tunnel!
Componentsedit* The software aspect of the system runs on standard home computer hardware and can be linked to other applications or databases! 6t first uses a series of image manipulation techni8ues to detect, normali.e and enhance the image of the number plate, and then optical character recognition (B&) to e"tract thealphanumerics of the license plate! #$%& systems are generally deployed in one of two basic approachesC one allows for the entire process to be performed at the lane location in real-time, and the other transmits all the images from many lanes to a remote computer location a nd performs the B& process there at some later point in time! When done at the lane site, the information captured of the plate alphanumeric, date-time, lane identification, and any other information re8uired is completed in appro"imately +3? milliseconds! This information can easily be transmitted to a remote computer for further processing if necessary, or stored at the lane for later retrieal! 6n the other arrangement, there are typically large numbers of %s used in a serer farm to handle high workloads, such as those found in the London congestion charge proAect! Bften in such systems, there is a re8uirement to forward images to the remote serer, and this can re8uire larger bandwidth transmission media!
Technology edit*
The font on1utch plates was changed to improe plate recognition!
#$%& uses optical character recognition (B&) on images taken by cameras! When1utch ehicle registration plates switched to a different style in +??+, one of the changes made was to the font, introducing small gaps in some letters (such as P andR ) to make them more distinct and therefore more legible to such systems! 'ome license plate arrangements use ariations in font si.es and positioning#$%& systems must be able to cope with such differences in order to be truly effectie! Eore complicated systems can cope with international ariants, though many programs are indiidually tailored to each country! The cameras used can include e"isting road-rule enforcement or closed-circuit teleision cameras, as well as mobile units, which are usually attached to ehicles! 'ome systems use i nfrared cameras to take a clearer image of the plates! 3*4*5*7*9* ?* *+**2*
ANPR in mobile systems edit*
The 1ubai police use #$%& cameras to monitor ehicles in front and either side of the patrol car
# Eerseyside %olice car e8uipped with mobile #$%&!
1uring the 99?s, significant adances in technology took automatic number plate recognition (#$%&) systems from limited e"pensie, hard to set up, fi"ed based applications to simple Gpoint and shootG mobile ones! This was made possible by the creation of software that ran on cheaper % based, non-specialist hardware that also no longer needed to be gien the pre-defined angles, direction, si.e and speed in which the plates would be passing the cameras field of iew! Further scaled-down components at more cost-effectie price points led to a record number of deployments by law enforcement agencies around the world! 'maller cameras with the ability to read license plates at higher speeds, along with smaller, more durable processors that fit in the trunks of police ehicles, allowed law enforcement officers to patrol daily with the benefit of license plate reading in real time, when they can interdict immediately! 1espite their effectieness, there are noteworthy challenges related with mobile #$%&s! Bne of the biggest is that the processor and the cameras must work fast enough to accommodate relatie
speeds of more than ?? mph (4? kmHh), a likely scenario in the case of oncoming traffic! This e8uipment must also be ery efficient since the power source is the ehicle battery, and e8uipment must be small to minimi.e the space it re8uires! &elatie speed is only one issue that affects the camera/s ability to actually read a license plate! #lgorithms must be able to compensate for all the ariables that can affect the #$%&/s ability to produce an accurate read, such as time of day, weather and angles between the cameras and the license plates! # system/s illumination waelengths can also hae a direct impact on the resolution and accuracy of a read in these conditions! 6nstalling #$%& cameras on law enforcement ehicles re8uires careful consideration of the Au"taposition of the cameras to the license plates they are to read! =sing the right number of cameras and positioning them accurately for optimal results can proe challenging, gien the arious missions and enironments at hand!
Algorithmsedit*
'teps +, and 2C The license plate is normali.ed for brightness and contrast, and then the characters are segmented to be ready for B&!
There are seen primary algorithms that the software re8uires for identifying a license plateC ! %late locali.ation @ responsible for finding and isolating the plate on the picture! +! %late orientation and si.ing @ compensates for the skew of the plate and adAusts the dimensions to the re8uired si.e! ! $ormali.ation @ adAusts the brightness and contrast of the image! 2! haracter segmentation @ finds the indiidual characters on the plates! 3! Bptical character recognition!
4! 'yntacticalHeometrical analysis @ check characters and positions against country-specific rules! 5! The aeraging of the recognised alue oer multiple fieldsHimages to produce a more reliable or confident result! 0specially since any single image may contain a reflected light flare, be partially obscured or other temporary effect! The comple"ity of each of these subsections of the program determines the accuracy of the system! 1uring the third phase (normali.ation), some systems use edge detection techni8ues to increase the picture difference between the letters and the plate backing! # median filter may also be used to reduce the isual noise on the image!
Difficultiesedit*
0arly #$%& systems were unable to read white or siler lettering on black background, as permitted on => ehicles built prior to 95!
'wedish licenseplate
Eust be able to recogni.e international license plates as such!
There are a number of possible difficulties that the software must be able to cope with! These includeC •
•
•
•
%oor file resolution, usually because the plate is too far away but sometimes resulting from the use of a low-8uality camera! :lurry images, particularly motion blur ! %oor lighting and low contrast due to oere"posure, reflectionor shadows! #n obAect obscuring (part of) the plate, 8uite often a tow bar, or dirt on the plate!
•
•
•
# different font, popular for anity plates (some countries do not allow such plates, eliminating the problem)! ircumention techni8ues! Lack of coordination between countries or states! Two cars from different countries or states can hae the same number but different design of the plate!
While some of these problems can be corrected within the software, it is primarily left to the hardware side of the system to work out solutions to these difficulties! 6ncreasing the height of the camera may aoid problems with obAects (such as other ehicles) obscuring the plate but introduces and increases other problems, such as the a dAusting for the increased skew of the plate! Bn some cars, tow bars may obscure one or two characters of the license plate! :ikes on bike racks can also obscure the number plate, though in some countries and Aurisdictions, such as Dictoria, #ustralia, Gbike platesG are supposed to be fitted! 'ome small-scale systems allow for some errors in the license plate! When used for giing specific ehicles access to a barricaded area, the decision may be made to hae an acceptable error rate of one character! This is because the likelihood of an unauthori.ed car haing such a similar license plate is seen as 8uite small!
Imaging hardwareedit* #t the front end of any #$%& system is the imaging hardware which captures the image of the license plates! The initial image capture forms a critically important part of the #$%& system which, in accordance to the arbage 6n, arbage But principle of computing, will often determine the oerall performance! License plate capture is typically performed by speciali.ed cameras designed specifically for the task, although new software techni8ues are being implemented that support any 6!%! based sureillance camera and increase the utility of #$%& for perimeter security applications! Factors which pose difficulty for license plate imaging cameras i nclude speed of the ehicles being recorded, arying ambient lighting conditions, headlight glare and harsh enironmental conditions! Eost dedicated license plate capture cameras will incorporate infrared illumination in order to sole the problems of lighting and plate reflectiity! Eany countries now use license plates that are retroreflectie!3* This returns the light back to the source and thus improes the contrast of the image! 6n some countries, the characters on the plate are not reflectie, giing a high leel of contrast with the reflectie background in any lighting conditions! # camera that makes use of actie infrared imaging (with a normal colour filter oer the lens and an infrared illuminator ne"t to it) benefits greatly from this as the infrared waes are reflected back from the plate! This is only possible on dedicated #$%& cameras, howeer, and so cameras used for other purposes must rely more heail y on the software capabilities! Further, when a full-colour image is re8uired as well as use of the #$%&-retrieed details it is necessary to hae one infrared-enabled camera and one normal (colour) camera working together! To aoid blurring it is ideal to hae the shutter speed of a dedicated camera set to H??? of a second! :ecause the car is moing, slower shutter speeds could result in an image which is too blurred to read using the B& software, especially if the camera is much higher up than the ehicle! 6n slow-moing traffic, or when the camera is at a lower leel and the ehicle is at an angle approaching the camera, the shutter speed does not need to be so fast! 'hutter speeds of H3?? of a second can cope with traffic moing up to 2? mph (42 kmHh) and H+3? of a second up to 3 mph (7 kmHh)! License plate capture cameras can now produce usable images from ehicles traeling at +? mph (9? kmHh)!
To ma"imi.e the chances of effectie license plate capture, installers should carefully consider the positioning of the camera relatie to the target capture area! 0"ceeding threshold angles of incidence between camera lens and license plate will greatly reduce the probability of obtaining usable images due to distortion! Eanufacturers hae deeloped tools to help eliminate errors from the physical installation of license plate capture cameras
Circumvention techniques edit* Dehicle owners hae used a ariety of techni8ues in an attempt to eade #$%& systems and roadrule enforcement cameras in general! Bne method increases the reflectie properties of the lettering and makes it more likely that the system will be unable to locate the plate or produce a high enough leel of contrast to be able to read it! This is typically done by using a plate coer or a spray, though claims regarding the effectieness of the latter are disputed! 6n most Aurisdictions, the coers are illegal and coered under e"isting laws, while in most countries there is no law to disallow the use of the sprays!4*5*Bther users hae attempted to smear their license pl ate with dirt or utili.e coers to mask the plate! $oelty frames around Te"as license plates were made illegal in Te"as on 'eptember +?? by Te"as 'enate :ill 29 because they caused problems with #$%& deices! That law made it a lass misdemeanor (punishable by a fine of up to =' I+??), or lass : (punishable by a fine of up to =' I+,??? and 7? days in Aail) if it can be proen that the owner did it to deliberately obscure their plates!7* The law was later clarified in +??5 to allow $oelty frames! 6f an #$%& 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! 6n order to aoid sureillance or penalty charges, there has been an upsurge in car cloning! This is usually achieed by copying registration plates from another car of a similar model and age! This can be difficult to detect, especially as cloners may change the registration plates and trael behaior to hinder inestigations!
Police enforcementedit*
Eobile #$%& cameras fitted to a$ew 'outh Wales %olice Force
losed-circuit teleision cameras such as these can be used to take the images scanned by automatic number plate recognition systems
Australiaedit* 'eeral 'tate %olice Forces, and the 1epartment of Justice (Dictoria) 9* use both fi"ed and mobile #$%& systems! The $ew 'outh Wales %olice Force
elgiumedit* The city of Eechelen uses an #$%& system since 'eptember +? to scan all cars crossing the city limits (inbound and outbound)! ars listed on /black lists/ (no insurance, stolen, etc!) generate an alarm in the dispatching room, so they can be intercepted by a patrol! #s of early +?+, million cars per week are automatically checked in this way!+*
!enmar"edit* The techni8ue is tested by the 1anish police! 6t will be in permanent use from the end of +?3! +2*
#ranceedit* 7? gantries oer maAor roads hae been built throughout the country! These together with a further +3? fi"ed cameras is to enable a ley of an eco ta" on lorries oer !3 tonnes! The system is currently being opposed and whilst they may be collecting data on ehicles passing the cameras, no eco ta" is being charged!+3*
$ermanyedit* Bn Earch +??7, the Federal onstitutional ourt of ermany ruled that some areas of the laws permitting the use of automated number plate recognition systems in ermany iolated the right topriacy!+4* Eore 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 iolation of erman law! These systems were proided by Jenoptik &obot mb<, and called Traffiapture!+5*
%ungaryedit* 'eeral
&ur"eyedit* 'eeral cities hae testedand some hae put into sericethe Gity 'ecurity #dministration 'ystemG, i!e!, capital #nkara, has debuted >K'- G>ent uenlik Konetim 'istemiG which consists of a registration plate number recognition system on the main arteries and city e"its! +9* The system has been used with two cameras per lane, one for plate recognition, one for speed detection! $ow the system has been widened to network all the registration number cameras together, and enforcing aerage speed oer preset distances! 'ome arteries hae 5?>mh limit, and some 3? kmh, and photo eidence with date-time details are posted to registration address if speed iolation is
detected! #s of +?+, the fine for e"ceeding the speed limit for more than ? is appro"imately ='I53!
"raineedit* The proAect of system integration MBLL6 TechnologyN and the Einistry of 6nternal #ffairs of =kraine1epartment of 'tate Traffic 6nspection ('T6) e"periments on the introduction of a modern technical comple" which is capable to locate stolen cars, driers depried of driing licenses and other problem cars in real time! The =krainian comple" GDideo controlG?* working by a principle of ideo fi"ing of the car with recognition of license plates with check under data base!
nited (ingdom edit*
#n #$%& 08uipped Dectra of thereater Eanchester %olice force
Main article: Police-enforced ANPR in the UK
The => has an e"tensie (#$%&) automatic number plate recognition TD network! 0ffectiely, the police and security serices track all car moements around the country and are able to track any car in close to real time! Dehicle moements are stored for + years in the $ational #$%& 1ata enter to be analy.ed for intelligence and to be used as eidence! 6n 995 a system of one hundred #$%& cameras, codenamed L=TTB$, was installed to feed into the automated :ritish Eilitary 6ntelligence 'ystems in $orthern 6reland!citation needed * Further cameras were also installed on the :ritish mainland, including unspecified ports on the east and west coasts! citation needed *
nited )tates edit*
# ity of #le"andria police car e8uipped with mobile #L%&!
#$%& cameras in operation on the:rooklyn :ridge in $ew Kork!
6n the =nited 'tates, #$%& systems are more commonly referred to as #L%& (#utomatic License %late &eaderH&ecognition) technology, due to differences in language (i!e!, Gnumber platesG are referred to as Glicense platesG in #merican 0nglish) Eobile #$%& use is widespread among =' law enforcement agencies at the city, county, state and federal leel! #ccording to a +?+ report by the %olice 0"ecutie &esearch Forum, appro"imately 5 of all =' police departments use some form of #$%&! * Eobile #$%& is becoming a significant component of municipal predictie policing strategies and intelligence gathering,+* as well as for recoery of stolen ehicles, identification of wanted felons, and reenue collection from indiiduals who are delin8uent on city or state ta"es or fines, or monitoring for G#mber #lertsG! 'uccessfully recogni.ed plates may be matched against databases including Gwanted personG, Gprotection orderG, missing person, gang member, known and suspected terrorist, superised release, immigration iolator, and $ational 'e" Bffender lists!* 6n addition to the real-time processing of license plate numbers, #L%& systems in the =' collect (and can indefinitely store) data from each license plate capture! 6mages, dates, times and %' coordinates can be stockpiled and can help place a suspect at a scene, aid in witness identification, pattern recognition or the tracking of indiiduals! #n early, priate sector mobile #$%& application has been applications for ehicle repossession and recoery2*), although the application of #$%& by priate companies to collect information from priately owned ehicles or collected from priate property (for e"ample, drieways) has become an issue of sensitiity and public debate! 3* Bther #L%& uses include parking enforcement, and reenue collection from indiiduals who are delin8uent on city or state ta"es or fines! The technology is often featured in the reality TD show Parking Wars featured on #O0 $etwork! 6n the show, tow truck driers and booting teams use the #L%& to find delin8uent ehicles with high amounts of unpaid parking fines!
)audi Arabiaedit* Dehicle registration plates in 'audi #rabia use white background, but seeral ehicle types may hae a different background! =nited 'tates diplomatic plates hae the letters /='1/, which in #rabic reads /1'=/ when read from right to left in the direction of #rabic script! There are only 5 #rabic letters used on the registration plates! 4* # hallenge for plates recognition in 'audi #rabia is the si.e of the digits! 'ome plates use both 0astern #rabic numerals and the /Western #rabic/ e8uialents! # research with source code is aailable for #%$& #rabic digits! 5*
)*edenedit* The techni8ue is tested by the 'wedish police at nine different places in 'weden!7*
Average-speed camerasedit* Main article: Speed limit enforcement
#$%& is used for speed limit enforcement in #ustralia, #ustria, 9* :elgium,2?* 1ubai (=#0),citation needed * France, 6taly,2* The $etherlands,2+* 'pain,2* and the =>!22* This works by tracking ehicles/ trael time between two fi"ed points, and calculating the aerage speed! These cameras are claimed to hae an adantage oer traditional speed cameras in maintaining steady legal speeds oer e"tended distances, rather than encouraging heay braking on approach to specific camera locations and subse8uent acceleration back to illegal speeds! 23*
+talyedit* 6n 6talian
&he Netherlandsedit* #erage speed cameras (trajectcontrole) are in place in the $etherlands since +??+! #s of July +??9, + cameras were operational, mostly in the west of the country and along the #+!23* 'ome of these are diided in seeral PsectionsQ to allow for cars leaing and entering the motorway! # first e"perimental system was tested on a short stretch of the #+ in 995 and was deemed a big success by the police, reducing oerspeeding to ?!44, compared to 3 to 4 when regular speed cameras were used at the same location! 25* The first permanent aerage speed cameras were installed on the # in +??+, shortly after the speed l imit was reduced to 7? kmHh to limit noise and air pollution in the area!27* 6n +??5, aerage speed cameras resulted in !5 million fines for oerspeeding out of a total of 9!5 millions! #ccording to the 1utch #ttorney eneral, the aerage number of iolation of the speed limits on motorway sections e8uipped with aerage speed cameras is between and +, compared to ? to 3 elsewhere! 29*
(edit* See also: Road speed limit enforcement in the United Kingdom
Bne of the most notable stretches of aerage speed cameras in the => is found on the #55 road in 'cotland, with + miles (3 km) being monitored between lasgow and #yr !3?* 6n +??4 it was confirmed that speeding tickets could potentially be aoided from the /'%0'/ cameras by changing lanes and the Foundation feared that people may play G&ussian &ouletteG changing from one lane to another to lessen their odds of being caught! 22*
Crime deterrentedit* &ecent innoations hae contributed to the adoption of #$%& for perimeter security and access control applications at goernment facilities! Within the =', Ghomeland securityG efforts to protect against alleged Gacts of terrorismG hae resulted in a doption of #$%& for sensitie facilities such as embassies, schools, airports, maritime ports, military and federal buildings, law enforcement and goernment facilities, and transportation centers! #$%& is marketed as able to be implemented through networks of 6% based sureillance cameras that perform Gdouble dutyG alongside facial recognition, obAect tracking, and recording systems for the purpose of monitoring suspicious or anomalous behaior, improing access control, and matching against watch lists! #$%& systems are most commonly installed at points of significant sensitiity, ingress or egress! EaAor =' agencies such as the 1epartment of
London and $ew Kork ity to proide citywide protection against acts of terrorism, and to proide support for public gatherings and public spaces!3*
Enterprise security and services edit* 6n addition to goernment facilities, many priate sector industries with facility security concerns are beginning to implement #$%& solutions! 0"amples include casinos, hospitals, museums, parking facilities, and resorts!32* 6n the =', priate facilities typically cannot access goernment or police watch lists, but may deelop and match against their own databases for customers, D6%s, critical personnel or Gbanned personG lists! 6n addition to proiding perimeter security, priate #$%& has serice applications for alet H recogni.ed customer and D6% recognition, logistics and key personnel tracking, sales and adertising, parking management, and logistics (endor and support ehicle tracking)!
Traffic controledit*
Dideo tolling at 'chRnberg, #ustria
Eany cities and districts hae deeloped traffic control systems to help monitor the moement and flow of ehicles around the road network! This had typically inoled looking at historical data, estimates, obserations and statistics, such asC •
ar park usage
•
%edestrian crossing usage
•
$umber of ehicles along a road
•
#reas of low and high congestion
•
Fre8uency, location and cause of road works
TD cameras can be used to help traffic control centres by giing them lie data, allo wing for traffic management decisions to be made in real-time! :y using #$%& on this footage it is possible to monitor the trael of indiidual ehicles, automatically proiding information about the speed and flow of arious routes! These details can highlight problem areas as and when they occur and help the centre to make informed incident management decisions! 'ome counties of the =nited >ingdom hae worked with 'iemens Traffic to deelop traffic monitoring systems for their own control centres and for the public! 33* %roAects such as
taken from TD cameras! #$%& systems can be used to proide aerage point-to-point Aourney times along particular routes, which can be displayed on a ariable-message sign(DE') giing driers the ability to plan their route! &BE#$'0 also allows traellers to see the current situation using a mobile deice with an 6nternet connection (such as W#%, %&' or ), allowing them to iew mobile deice TD images within the company Trafficmaster has used #$%& since 997 to estimate aerage traffic speeds on non-motorway roads without the results being skewed by local fluctuations caused by traffic lights and similar! The company now operates a network of oer 2??? #$%& cameras, but claims that only the four most central digits are identified, and no numberplate data is retained! 34*35*37* •
6000 transactions on 6ntelligent Transportation 'ystems(6000 6ntelligent Transportation 'ystems 'ociety) published some papers on the plate number recognition technologies and applications!relevant ! disc"ss*
Electronic toll collection edit* &oll roads edit*
The FasTrak system in Brange ounty uses #$%& and radiotransponders
Film showing the approach to and passing of a toll station in 6taly, using a Telepass B:=! $ote the yellow Telepass lane signs and road markings and the sound emitted by the B:= when passing the lane
Bntario/s 2?5 0T& highway uses a combination of #$%& and radiotransponders to toll ehicles entering and e"iting the road! &adio antennas are located at each Aunction and detect the transponders, logging the uni8ue identity of each ehicle in much the same way as the #$%& system does! Without #$%& as a second system it would not be possible to monitor all the traffic! 1riers who opt to rent a transponder for I+!33 per month are not charged the GDideo Toll hargeG of I!4? for using the road, with heay ehicles (those with a gross weight of oer 3,??? kg) being re8uired to use one! =sing either system, users of the highway are notified of t he usage charges by post! There are numerous other electronic toll collection networks which use this combination of &adio fre8uency identification and #$%&! These includeC
•
•
•
•
The olden ate :ridge in 'an Francisco, alifornia, began using an all-electronic tolling system combining Fastrak and #$%& on Earch +5, +?!39* $ Suick %ass for the 6nterstate 32? ($orth arolina) Triangle 0"pressway in Wake ounty, $orth arolina :ridge %ass4?* for the 'aint John
•
ityLink O 0astlink in Eelbourne, #ustralia
•
ateway Eotorway and Logan Eotorway, :risbane, #ustralia
•
FasTrak in alifornia, =nited 'tates
•
•
Tunnels in ong
•
•
#utopista entral4+* in 'antiago, hile (site in 'panish) 0-%ass in $ew Kork, $ew Jersey, Eassachusetts (as Fast Lane until +?+), Dirginia (formerly 'mart Tag), and other states! Earyland &oute +?? uses a combination of 0-%ass and #$%&!
•
TollTag in $orth Te"as!
•
6-%ass in 6llinois
•
%ike %ass in Bklahoma
•
%each %ass 6-73 #tlanta, # winnett ounty
•
B' (Btomatik eUiV 'istemi) used at :osphorus :ridge, Fatih 'ultan Eehmet :ridge, and Trans 0uropean Eotorway entry points in stanbul, Turkey
•
E3? Westlink Toll in 1ublin, 6reland
•
orea
•
$orthern ateway, '< , #uckland, $ew ealand
•
•
0ergreen %oint Floating :ridge, 'eattle, and Washington 'tate &oute 45
Portugal edit* This section does not cite any references or sources . Please help improve this section by adding citations to reliable sources . Unsourced material may be challenged and removed. (December 2012)
%ortuguese roads hae old highways with toll station where driers can pay with cards and also lanes where there are electronic collection systems!
Charge ,ones the London congestion chargeedit*
The London congestion chargescheme uses +? cameras and #$%& to help monitor ehicles in the charging .one
The London congestion charge is an e"ample of a system that charges motorists entering a payment area! Transport for London(TfL) uses #$%& systems and charges motorists a daily fee of X? paid before ?pm if they enter, leae or moe around within the congestion charge .one between 5 a!m! and 4C?? p!m!, Eonday to Friday! # reduced fee of X9 is paid by ehicle owners who sign up for the automatic deduction scheme! Fines for traeling within the .one without paying the charge are X4? per infraction if paid before the deadline, doubling to X+? per infraction thereafter! There are currently ,3?? cameras which use automatic number plate recognition (#$%&) technology!42* There are also a number of mobile camera units which may be deployed anywhere in the .one! 6t is estimated that around 97 of ehicles moing within the .one are caught on camera! The ideo streams are transmitted to a data centre located in central London where the #$%& software deduces the registration plate of the ehicle! # second data centre proides a backup location for image data!
:oth front and back number plates are being captured, on ehicles going both in and out @ this gies up to four chances to capture the number plates of a ehicle entering and e"iting the .one! This list is then compared with a list of cars whose ownersHoperators hae paid to enter the .one @ those that hae not paid are fined! The registered owner of such a ehicle is looked up in a database proided by the 1DL#! 43*
)*edenedit* 6n 'tockholm, 'weden, #$%& is used for the 'tockholm congestion ta", owners of cars driing into or out of the inner city must pay a charge, depending on the time of the day! From +?, also for the othenburg congestion ta", which also includes ehicles passing the city on the main highways!
sageedit* 'eeral => companies and agencies use #$%& systems! These include Dehicle and Bperator 'erices #gency (DB'#), 44* 1rier and Dehicle Licensing #gency (1DL#)45* and Transport for London!47*
Challengesedit* Controversyedit* The introduction of #$%& systems has led to fears of misidentification and the furthering of #$%&style sureillance!49* 6n the =nited 'tates, some such as regg 0asterbrook oppose what they call Gmachines that issue speeding tickets and red-light ticketsG as the beginning of a slippery slope towards an automated Austice systemC G# machine classifies a person as an offender, and you can/t confront your accuser because there is no accuser!!! can it be wise to establish a principle that when a machine says you did something illegal, you are presumed guiltyYG5?* 'imilar criticisms hae been raised in other countries! 0asterbrook also argues that this technology is employed to ma"imi.e reenue for the state, rather than to promote safety!5?* The electronic sureillance system produces tickets which in the =' are often in e"cess of I??, and are irtually impossible for a citi.en to contest in court without the help of an attorney!citation needed * The reenues generated by these machines are shared generously with the priate corporation that builds and operates them, creating a strong incentie to tweak the system to generate as many tickets as possible! Blder systems had been notably unreliable; in the => this has been known to lead to charges being made incorrectly with the ehicle owner haing to pay X? in order to be issued with proof (or not) of the offense! 6mproements in technology hae drastically decreased error rates, but false accusations are still fre8uent enough to be a problem! %erhaps the best known incident inoling the abuse of an #$%& database in $orth #merica is the case of 'dmonton S"n reporter >erry 1iotte in +??2! 1iotte wrote an article critical of 0dmonton police use of traffic cameras for reenue enhancement, and in retaliation was added to an #$%& database of Ghigh-risk driersG in an attempt to monitor his habits and create an opportunity to arrest him! 5*5+*5* The police chief and seeral officers were fired as a result, and The Bffice
of the %riacy ommissioner of anada e"pressed public concern oer the Ggrowing police use of technology to spy on motorists!G 52* Bther concerns include the storage of information that could be used to identify people and store details about their driing habits and daily life, contraening the 1ata %rotection #ct along with similar legislation (see personally identifiable information)! The laws in the => are strict for any system that uses TD footage and can identify indiiduals! 53* 53*54*55*57*59*7?*7*7+*
#lso of concern is the safety of the data once it is mined, following the discoery of police sureillance records lost in a gutter!7*72* There is also a case in the => for saying that use of #$%& cameras is against the law under the&egulation of 6nestigatory %owers #ct +???! 73* The breach e"ists, some say, in the fact that #$%& is used to monitor the actiities of law-abiding citi.ens and treats eeryone like the suspected criminals intended to be sureyed under the act! The police themseles hae been known to refer to the system of #$%& as a G+2H5 traffic moement databaseG which is a diersion from its intended purpose of identifying ehicles inoled in criminal actiities! 74* The opposing iewpoint is that where the plates hae been cloned, a /read/ of an innocent motorist/s ehicle will allow the elimination of that ehicle from an inestigation by isual e"amination of the images stored! Likewise, stolen ehicles are read by #$%& systems between the time of theft and report to the %olice, assisting in the inestigation! The Associated Press reported in #ugust +? that $ew Kork %olice 1epartment cars and license plate tracking e8uipment purchased with federal <61T# (
Plate inconsistency and .urisdictional differences edit*
Eany #$%& systems claim accuracy when trained to match plates from a single Aurisdiction or region, but can fail when trying to recogni.e plates from other Aurisdictions due to ariations in format, font, color, layout, and other plate features!9* 'ome Aurisdictions offer anity or affinity plates (particularly in the ='), which can create many ariations within a single Aurisdiction!9+* From time to time, =' states will make significant changes in their license plate protocol that will affect B& accuracy! They may add a character or add a new license plate design! #L%& systems must adapt to these changes 8uickly in order to be effectie! #nother challenge with #L%& systems is that some states hae the same license plate protocol! For e"ample more than one state uses the standard three letters followed by four numbers! 'o each time the #L%& systems alarms, it is the userZs 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 ma"imum effectieness, an #$%& system should be able to recogni.e plates from any Aurisdiction, and the Aurisdiction to which they are associated, but these many ariables make such tasks difficult! urrently at least one =' #$%& proider (%late'mart) claims their system has been independently reiewed as able to accurately recogni.e the =' state Aurisdiction of license plates, and one 0uropean #$%& proider claims their system can differentiate all 0= plate Aurisdictions!9*92*
Accuracy and measurement of ANPR system performanceedit* # +??7 article in Parking (rend )nternational discussed a disparity in claimed s! e"perienced license plate recognition read rates, with manufacturers claiming that their recognition engines can correctly report 97 of the time, although customers e"perience only 9? to 92 success, een with new e8uipment under perfect conditions! 0arly systems were reportedly only 4? to 7? reliable! 93* True system error rate is the product of its subsystem error rates (image capture, license plate image e"traction, L% image interpretation); slight increases in subsystem error rates can produce dramatic reductions of read rates! The effects of real-world interfering factors on read rate are not uniformly specified or tested by manufacturers! The article states Gthere is a need for the industry to adopt a standard performance measurement protocol to enable potential customers assess the best fit for their particular re8uirements!G93*
!ther usesedit* #$%& systems may also be used forHbyC •
•
'ection control, to measure aerage ehicle speed oer longer distances!94* :order crossings
•
•
•
•
•
•
•
•
#utomobile repossessions2*95* petrol stations to log when a motorist dries away without paying for their fuel! # marketing tool to log patterns of use Targeted adertising, a-la GEinority &eportG-style billboards!97*99* Traffic management systems, which determine traffic flow using the time it takes ehicles to pass two #$%& sites??* #nalyses of trael behaiour (route choice, origin-destination etc!) for transport planning purposes ?*?+* 1rie Through ustomer &ecognition, to automatically recogni.e customers based on their license plate and offer them the items they ordered the last time they used the serice! To assist isitor management systems in recogni.ing guest ehicles!
•
%olice and #u"iliary %olice
•
ar parking companies
•
"elated research society edit* •
6000 6ntelligent Transportation 'ystems 'ociety
#ee alsoedit* Wikimedia Commons has media related to Automatic number plate recognition .
•
#6 effect
•
#pplications of artificial intelligence
•
losed circuit teleision
•
Facial recognition system
•
%arking lot
•
&oad %olicing =nit
•
'%0' (speed camera)
•
Dehicle location data
Lists •
List of emerging technologies
•
Butline of artificial intelligence
"eferencesedit* ! /ump up0 G#$%& TutorialG! #$%& Tutorial! 3 #ugust +??4! &etrieed +?+-?-+2! +! /ump up0 G'han 1u ; 6ntelliDiew Technol!, 6nc!, algary, #:, anada ; 6brahim, E! ; 'hehata, E! ; :adawy, Wael; #utomatic License %late &ecognition (#L%&)C # 'tate-of-the-#rt &eiewG! 6000! Feb +?! &etrieed +?2-?-?9! ! /ump up0 G#n introduction to #$%&G! ct-information!co!uk! &etrieed +?+-?-+2! 2! /ump up0 GTD network tracks /getaway/ carG! :: $ews! + $oember +??3! &etrieed +?-?7-+! 3! /ump up0 G%late &ecognitionG! %hotoop!com! 4! /ump up0 G#lgorithm for License %late &ecognitionG! D6'L, Technion! +??+! 5! /ump up0 G# &eal-time ehicle License %late &ecognition (L%&)G! D6'L, Technion, +?? 7! /ump up0 G#n #pproach To License %late &ecognitionG(%1F)! =niersity of algary! 994! &etrieed+?+?-+2! 9! /ump up0 1raghici, 'orin (995)! G# neural network based artificial ision system for license plate recognitionG (%1F)! 1ept! of omputer 'cience,Wayne 'tate =niersity! &etrieed +?+?-+2! ?! /ump up0 GLicense %late &ecognition in Turkey (%laka Bkuma 'istemi)G! rimedia!com! &etrieed+?+-?-+2!
! /ump up0 >wa[nicka, ahraman, Fatih; okmen, Euhittin (+??)! GLicense %late haracter 'egmentation :ased on the abor Transform and Dector Suanti.ationG(%1F)! #rchied from the original on +??4-?3-+2! &etrieed +?+-?-+2! ! /ump up0 BndreA Eartinsky (+??5)! G#lgorithmic and mathematical principles of automatic number plate recognition systemsG (%1F)! :rno =niersity of Technology! &etrieed +?+?-+2! 2! /ump up0 G# License %late &ecognition algorithm for 6ntelligent Transportation 'ystem applicationsG!=niersity of the #egean and $ational Technical =niersity of #thens! +??4! #rchied from the original on +??7-?2-+?! &etrieed +?+-?-+2! 3! /ump up0 G#utomatic $umber %late &ecognitionG! ctinformation!co!uk! &etrieed +?+-?-+2! 4! /ump up0 'e"ton, 'tee ( July +??)! GLicense-plate spray foils traffic camerasG! Washington (imes! &etrieed +?-?5-?+! 5! /ump up0 'e"ton, 'tee ( July +??)! GLicense-plate spray foils traffic camerasG! Washington (imes! &etrieed +?-?5-?+! 7! /ump up0 Wentworth, Jeff ! GBbscured license plate could be motorists/ ticket to fineG! Te"as 'tate 'enate! #rchied from the original on +2 Earch +??3! &etrieed 3 #pril +??3! 9! /ump up0 *
dead link *
+?! /ump up0 GeTendering - ontract #ward $otice 1etail Diew ??9+??&FTG! Tenders!nsw!go!au! &etrieed+?2-?2-+! +! /ump up0httpCHHwww!police!nsw!go!auH\\dataHassetsHpdf\fileH???7H72 774H#nnual\&eport\-\Kear\in\&eiew!pdf ++! /ump up0
+4! /ump up0 G1as :undeserfassungsgerichtG (in erman)! :erfg!de! $oember +??7! &etrieed+??9-?+-4! +5! /ump up0 GTraffiaptureG! Jenoptic!com! +7! /ump up0 GEatri" %oliceG! en (+7 February +??)! GThe Wired &epo EanC
3! /ump up0 GLicense plate data not Aust for copsC %riate companies are tracking your carG! N0 Ne.s! 9 July +?! 4! /ump up0 GDehicle registration plates of 'audi #rabiaG! 5! /ump up0 G#rabic Licence %late &ecognitionG! 'ourceforge! 7! /ump up0 G%olisens nya apenC GTerminatorG-teknikG! 'wedenC 0"pressen!se! 9! /ump up0 G'ection ontrolC +2!??? &aser ange.eigtG (in erman)! #ustriaC B&F!at! 5 June +??! 2?! /ump up0 GFlitsers!netG! 2! /ump up0 G0cco come fun.iona il Tutor in autostrada - $oti.ie brei - $ews - 6nforma.ione - #!'!#!%!'! 6l %ortale della 'icure..a 'tradaleG! #saps!it! &etrieed +?+-?7-?! 2+! /ump up0 G1escription of the systemG (in 1utch)! 1utch #ttorney eneral! 1ecember +??9! &etrieed+?+-?-+2! 2! /ump up0 GLlegan los radares de elocidad media a 0spa`aG (in 'panish)! $oticias!coches!com! + 'eptember +??9! &etrieed +?+-?-+2!
22! ^ Jump up toC G'peeding tickets can potentially be aoided by changing lanesG! (he 1ail2 Mail (London)! 3 Bctober +??4! &etrieed +?+-?-+2! The
23! ^ Jump up toC GFre8uently #sked Suestion oer TraAectcontroleG (in 1utch)! 1utch #ttorney eneral! &etrieed +?+-?-+2! a b
24! /ump up0 G#utostrade '!p!#G (in 6talian)! #utostrade!it! +??7-?3+3! &etrieed +?+-?7-?! 25! /ump up0 G$iemand riAdt meer te hard op de #+G (in 1utch)! (ro". ! 7 Bctober 995! &etrieed+?+-?-+2! 27! /ump up0 G$ummerborden le.en op de #G (in 1utch)! (ro". ! 4 Eay +??+! &etrieed +?+-?-+2! 29! /ump up0 >reling, Tom (+7 #ugust +??7)! G0en duur foutAe an de computerG (in 1utch)! NR0 3andels4lad ! &etrieed +?+-?+2! 3?! ^ Jump up toC G'%0' - aerage speed cameras on '%0'G! #55safetygroup!com! 2 July +??3! &etrieed +?+-?-+2! a b
3! /ump up0 GJeremy larkson tilts at windmills @ 'peed camera aoidance is an urban mythG! (he Register ! + July +??7! &etrieed + January+?+! 3+! /ump up0 base-security-and-inestigationsH G=' $ay 'elects %late'mart 0"clusiely to
37! /ump up0 GTrunk &oads @ %TFEG! Trafficmaster!co!uk! * &etrieed +?+-?-+2! dead link
39! /ump up0 eeping 972 in the pastG! (he 7"ardian (=>)! 5?! ^ Jump up toC 0asterbrook, regg (+7 February +??3)!G1aily 0"press - Lights, amera, #ctionG! (he Ne. Rep"4lic *nline ! #rchied from the originalon +??3-?9-?!(subscription re8uired) a b
5! /ump up0 G0dmonton officer reprimanded for checking up on newspaper columnistG! : $ews! 2 $oember +??3! 5+! /ump up0 G: 0dmonton - Features - 0dmonton %oliceG! : $ews! #rchied from the originalon +??5-?-+! &etrieed + July +?! 5! /ump up0 GBfficer faces suspension for Bertime stingG ! 0dmonton %olice Watch! 4 'eptember +??7! &etrieed +?+-?+2! 52! /ump up0 GanadaC %riacy ommissioner oncerned Ber License %late 'pyingG! Thenewspaper!com! July +??5! &etrieed +?+-?-+2!
53! ^ Jump up toC Lettice, John (+ February +??)! GThe London charge .one, the 1% #ct, and E' !$0TG! LondonC The &egister! &etrieed +?+-?-+2! a b
54! /ump up0 G#$%& 'trategy for the %olice 'erice +??3H+??4G! #%B (#ssociation of hief %olice Bfficers) #$%& 'teering roup! 5 Earch +??3! #rchied from the original on +??-?-! &etrieed +?+-?-+2! 55! /ump up0 G1riing crime downG (%1F)! LondonC => oernment number plate cameras go nationalG! The &egister! #rchied from the original on +7 Earch +??3! &etrieed +7 Earch +??3! 7+! /ump up0 G&ecogni.ing a new way to keep traffic moingG! 'iemens Traffic! #rchied from the original on 5 #pril +??3! &etrieed #pril +??3! 7! /ump up0 G=>C Traffic amera 1ata 1ropped in utterG! thenewspaper!com! &etrieed +?+-?-+2! 72! /ump up0 lements, Jon ( Earch +??7)! G# police memory stick containing secret data is found in gutterG ! (he Mirror 8UK9 ! &etrieed +?-?5-?+! 73! /ump up0 G&egulation of 6nestigatory %owers #ct +???G!legislation5gov5"k ! The $ational #rchies! +???! &etrieed +?+-?-+2! 74! /ump up0 G%olice number plate cameras may breach &6%# ommissionerG! The&egister!com! 5 July +??4! &etrieed +?+?-+2! 75! /ump up0 GEoney for nothingC the drug war and the war on EuslimsG! 77! /ump up0 reen, Jonah (+2 February +?+)! G$K%1 1efends Tactics Ber Eos8ue 'pying; &ecords &eeal $ew 1etails Bn Euslim 'ureillanceG!3"ffington Post !
79! /ump up0 G#L= raises priacy concerns about police technology tracking driers - $$!comG! 0NN ! 7 July +?! 9?! /ump up0 >lein, #llison (+4 'eptember +?+)! G#L= files suit oer info on license plate readersG! (he Washington Post ! 9! /ump up0 G#utomated License %late &ecognition 'ystemsC %olicy and Bperational uidance for Law 0nforcementG! U5S5 1epartment of 6"stice National )nstit"te of 6"stice! +?+! 9+! /ump up0 G$ew 1eelopments 6n #L%&G! Police (echnolog2 Maga;ine! 3 February +?! 9! /ump up0 +*
dead link *
92! /ump up0 the license plate recognition company (+??2-?+?)! GLicense %late &ecognition #lgorithms and TechnologyG! %laterecognition!info! &etrieed+?2-?2-+! 93! ^ Jump up toC GEeasuring #$%& 'ystem %erformanceG(%1F)! Parking (rend )nternational ! June +??7! &etrieed +?+-?-+2! a b
94! /ump up0 G'ection controlG! Derkeershandhaing 1ossiers! #rchied from the original on +??4-?3-+! &etrieed +?+-?-+2! 95! /ump up0 G :illboards 08uipped with License %late 'py amerasG! The$ewspaper!com! +3 'eptember +??9! &etrieed +?-?5-?+! 99! /ump up0 G=> :illboards 08uipped with License %late 'py amerasG! +3 'eptember +??9! &etrieed+?-?5-?+! ??! /ump up 0 G0"treme TD #nnounces ontract for 'tockholm Traffic amerasG (%1F)! 0"tremecct!com! ++ #pril +??2! #rchied from the original on +??7-?2-?9! &etrieed +?+-?-+2! ?! /ump up 0 Friedrich, Earkus; Jehlicka, %rokop; 'chlaich, Johannes (+??7)! G#utomatic number plate recognition for the obserance of trael behaiorG! =niersitt 'tuttgart 6nstitut fr 'traen und Derkehrswesen! &etrieed+?-?5-?+! ?+! /ump up 0 Friedrich, Earkus; Jehlicka, %rokop; 'chlaich, Johannes (+??7)! G7th 6nternational onference on 'urey Eethods in TransportC
ategoriesC
• • • • • • • • • •
#utomatic number plate recognition 'ureillance #pplications of computer ision #rtificial intelligence applications #uthentication methods 0lectronic toll collection Law enforcement e8uipment &oad traffic management Bptical character recognition #utomatic identification and data capture
$aigation menu •
reate account
•
Log in • •
•
&ead
•
0dit
•
Diew history o • • • • • • •
Eain page ontents Featured content urrent eents &andom article 1onate to Wikipedia Wikimedia 'hop 6nteraction
• • • • •
• • • • • • • •
•
#rticle Talk