License Plate Recognition for Parking Enforcement

License Plate Recognition (LPR): The Foundation of Modern Parking Enforcement

License plate recognition (LPR) is the use of a camera and optical character recognition software to turn a vehicle’s license plate into text that a system can match against its own records. In parking operations those records are permits, violation history, dwell time and vehicle watchlists.

Most parking professionals have heard of LPR. The wider gap is between knowing what it is and knowing what it takes to make it work in day-to-day enforcement: read accuracy under real conditions, what the camera is connected to, which deployment type suits the site, what a privacy review has to cover, and whether the plate read reaches the officer in time to matter.

This page works through that full picture. It is written for the person who has to choose a system and live with it, not for someone shopping for a camera.

A note on terminology: LPR and ALPR (automatic license plate recognition) name the same technology, and ANPR (automatic number plate recognition) is the term used in much of Europe and Asia. Canadian enforcement and government material tends to use ALPR; the parking industry tends to use LPR. Everything on this page applies equally to ALPR cameras, ALPR systems and ANPR systems, whichever acronym your team uses.

What is license plate recognition used for in parking?

A plate read is only useful if it answers a question an officer or an administrator actually has. In parking operations there are four: does this vehicle hold a valid permit, has it overstayed a time limit, does it carry unpaid violations, and is it on a watchlist. A connected system answers all four from the same read, in the same pass.

That is why LPR shows up across so many parts of a parking operation:

Permit validation. The plate is matched against live permit records, so an officer knows whether the vehicle is entitled to be where it is.
Time-based enforcement. Repeated reads of the same zone record when a vehicle was present, which is how dwell time is measured in digital tire chalking.
Violation history. The same read surfaces outstanding violations and repeat behaviour, so enforcement is consistent rather than opportunistic.
Security watchlists. Reads are checked against vehicle records that security staff are watching, independently of whether the vehicle is parked legally.
Access and gateless entry. Fixed cameras at entry and exit points identify vehicles without a gate, ticket or transponder.
Occupancy and turnover data. Entry and exit reads produce arrival and departure patterns that feed parking analytics.

How does license plate recognition work?

At the system level, LPR is an image recognition problem followed by a database lookup. A camera captures an image of a vehicle, software locates the plate within that image, reads the characters, and the resulting text string is compared against configured records.

The stages are the same in every system:

Image capture. A handheld, vehicle-mounted or fixed camera photographs the vehicle. Image quality depends on resolution, lighting, angle and the distance between camera and plate.
Plate detection. Software isolates the plate from the rest of the image: vehicle body, background, shadows and other visual noise.
Character recognition. OCR reads the characters, accounting for fonts, plate damage, dirt, and regional plate formats.
Database matching. The recognized string is compared against configured data, and a status is returned.
Action and record. Depending on the result the system confirms a permit, flags the vehicle for review, raises a watchlist alert, or logs the read for dwell-time tracking.

Where the read actually happens

On OPSCOM’s supported hardware, the camera reads the plate on board and transmits the plate text with its metadata. The server receives the read, not raw video. That matters for two reasons: a site does not need the bandwidth of a video pipeline, and the camera keeps working as a reader even though every part of the decision-making lives in the platform.

The server’s job is validation rather than reading. It normalizes reads from different camera vendors into one format, applies plate-lookup corrections, suppresses duplicate sightings inside a short window so the same vehicle passing two cameras is not counted twice, validates the plate against live permit records, raises watchlist alarms, and stores the event with its image. The image is uploaded as the read happens and is available as evidence on the record.

Because the lookup happens while the officer is still in the lot, there is no post-patrol reconciliation step. The plate has already been matched to the permit, the violation history and the watchlist before the pass is finished.

What affects read accuracy, and how is it handled?

Accuracy is the most quoted and least comparable number in the LPR market. Accuracy figures are measured in different conditions, against different plate populations, and with different definitions of what counts as a correct read, so a headline percentage from one supplier rarely predicts performance in your lots. OPSCOM does not publish a read-rate percentage. What it publishes instead is the setup that produces good reads and the behaviour of the system when a read is uncertain.

The factors that move accuracy in practice are the physical ones:

Plate condition. Dirty, damaged, faded or partially covered plates are harder to read, and a badly obscured plate may not be readable at all.
Angle and distance. Every camera has an optical range and a workable viewing angle. Plates outside that envelope produce weaker images and more errors.
Lighting. Direct sun, deep shadow, headlight glare and night conditions all change image quality, so exposure handling matters more than resolution claims.
Regional plate formats. North America carries hundreds of plate designs and fonts. A system not tuned for the plates in front of it will misread them.
Movement. Motion blur affects fast-moving vehicles. Parking enforcement is comparatively forgiving because vehicles are stationary or slow, which is why the same technology behaves differently in tolling than it does in a lot.

Two setup decisions matter more than any accuracy claim. The first is geometry: a fixed camera reads plates best with roughly 1.5 to 2.5 metres between camera and plate, while the mobile Survision camera is positioned 3 to 8 metres from the reference plate and less than 30 degrees off it. The second is the confidence threshold. The handheld workflow includes a configurable NOREAD setting, which decides what share of lower-confidence reads are shown to the officer, and OPSCOM’s setup guidance starts it at 50% and tunes from there to suit the site.

What separates a usable system from a frustrating one is what it does with the reads it is unsure about:

Fuzzy matching compensates for characters that LPR cameras commonly confuse (8, 3 and B, or 0, D, O and Q), so a plausible misread still finds the right record.
Plate lookup corrections rewrite plates the system consistently misreads at a given site, so a known pattern stops repeating.
Human corrections feed back into the read, so an officer’s fix improves the next pass rather than being lost.

The limits are worth stating plainly: OPSCOM’s own documentation notes that a perfect raw algorithmic read score is not achievable with standard LPR cameras. That is not a defect to be engineered away. It is the reason a low-confidence read should be flagged for a person rather than silently treated as accurate. An uncertain read presented as certain creates enforcement errors; the same read flagged for review creates a workflow.

Jason Dulmage, Systems Administrator at Fleming College, described the change after switching to OPSCOM’s PL8RDR platform: “Read accuracy, range of read angles and distance are significantly improved. The ability to mount this camera almost haphazardly and get scanning compared to previous troublesome mounting arrangements has been a real time savings.”

Which deployment type fits your operation?

LPR is deployed in three configurations. They differ in coverage, cost and how much infrastructure a site has to absorb, and most operations end up using more than one.

Handheld LPR

An officer uses a handheld device or tablet with a camera to scan plates directly. The read comes back against live records while the officer is standing there. Nothing is installed and no vehicle is needed.

Handheld LPR is the lowest-cost way to start and the most flexible: it works in lots, garages, on-street spaces and anywhere an officer can walk. The tradeoff is coverage, because an officer can only scan what they physically reach.

Best for: smaller operations, sites where officers already patrol on foot, selective validation rather than blanket coverage, and organisations that want to prove the workflow before investing in cameras.

Vehicle-mounted LPR

Cameras are mounted on a patrol vehicle and connected to an in-vehicle computer. As the vehicle drives the route, every plate in range is read without the officer stopping or touching anything. Mobile LPR is supplied as a complete kit: camera, tablet, printer and the PL8RDR computer.

Coverage stops depending on how many officers are walking. A single pass covers a whole route, and the system flags the vehicles that need a second look so the officer’s attention goes to exceptions rather than to checking every permit by hand.

Best for: large lots and portfolios, municipal enforcement across several zones, and operations where consistent coverage of the whole area is the goal.

Fixed camera LPR

Cameras are installed at specific points, such as entrances, exits, ramps and perimeter locations, and monitor continuously. Fixed cameras are what make gateless parking work: the plate is the credential, so there is no gate, ticket or transponder.

Fixed installations are usually the organisation’s own cameras, registered in the platform by name, IP address and type, and given an OPSCOM camera ID. Fixed cameras also generate the entry and exit records that entry and exit reads depend on for occupancy and turnover reporting.

Best for: controlled or gateless access, continuous coverage at a specific point, perimeter and security monitoring, and locations where every vehicle has to be accounted for.

OPSCOM’s PL8RDR platform supports all three modes in one system, so an operation can start handheld and add mobile or fixed reading without changing platform or rebuilding its workflows. Reads are recorded in a single audit log by event type: entry, exit, patrol, tracking or any.

What camera hardware does OPSCOM support?

OPSCOM does not manufacture cameras, and that is a deliberate position rather than a gap. PL8RDR reads from Survision, Tattile and Vaxtor-based cameras, and because the platform normalizes reads from each onto the same pipeline, the hardware decision stays a purchasing decision instead of a platform decision.

Survision: the PicoPak camera supplied in OPSCOM’s mobile LPR kit, connected to the in-vehicle PL8RDR computer.
Tattile: supported in both mobile and fixed installations.
Vaxtor (VaxALPR On Camera): fixed cameras running Vaxtor’s on-camera software, which includes Panasonic, Axis, i-PRO, Genetec and Hanwha. Supporting Vaxtor means supporting any camera that runs it, since the ingest path on OPSCOM’s side is the same.

The supply model follows the deployment type. Mobile LPR arrives as an OPSCOM kit with the camera included. Fixed cameras are typically the organisation’s own, installed on its network and integrated with the platform; organisations can source that hardware independently or buy it through OPSCOM.

Where a camera collects make, model and colour in addition to the plate, OPSCOM can carry that detail through to the read, enabled per client.

Why does integration decide whether LPR is worth it?

This is the part of an LPR decision that gets the least attention and changes the most.

A standalone LPR system collects plates. If those plates land in their own database, disconnected from permits and enforcement, someone still has to reconcile them by hand before anything can be done: officers finish a patrol with a list of reads that has to be matched against permit data before a single citation is defensible. The speed LPR creates in the field is handed back downstream.

Connected LPR removes that step. In OPSCOM, every read from PL8RDR is matched as it happens against:

ParkAdmin permit records: is this vehicle’s permit active, expired or suspended?
ViolationAdmin enforcement records: does this vehicle carry outstanding violations, and is it a repeat offender?
Digital chalking records: has this vehicle already been observed in this zone, and when?
IncidentAdmin watchlists: is this vehicle flagged for security attention or linked to an open investigation?

The difference between a standalone reader and a connected platform is not incremental. It is the difference between collecting plate data and acting on it in the same pass. The wider argument for one shared record of parking and security operations sits in the parking management systems and parking enforcement systems articles.

How does LPR work with digital tire chalking?

Time-based enforcement is the clearest example of why the connection matters more than the camera. Digital chalking tracks how long a vehicle has been present in a time-limited zone, and LPR supplies the plate identification that makes that tracking accurate at scale.

The two work from the same database rather than from separate tools. A patrol pass reads every plate in the zone and records each vehicle’s presence. A later pass reads the same zone again and the earlier observation is already there to compare against, so a vehicle that has overstayed is flagged with both observations attached as evidence, and the officer decides what happens next.

That is how Forks North Portage enforces time limits across 20+ downtown Winnipeg lots: vehicle-mounted LPR covering a dispersed portfolio that manual chalking could not cover consistently. The Forks North Portage case study shows what that looks like in practice.

Can LPR support security operations, not just parking?

In a parking-only system, a plate read answers a permit question. In a platform that also runs security incidents, the same read does more: it is checked against watchlists and vehicle records tied to open investigations, and the relevant staff are alerted when a vehicle of interest enters a monitored area. The vehicle’s parking status is not the trigger.

That capability matters most where a large site has to balance access with safety:

University campuses, where trespass orders, safety concerns or active investigations may require monitoring specific vehicles across a wide area.
Healthcare campuses, where patient safety and visitor management create legitimate needs beyond parking compliance.
Municipal operations, where enforcement and police liaison work from different mandates on the same streets.
Mixed-use developments, where the same vehicle repeatedly accessing residential areas needs to be identified consistently.

Carleton University runs PL8RDR connected to IncidentAdmin across a large urban campus in Ottawa, which makes it a good example of LPR operating as a campus safety tool rather than only a parking enforcement tool. The Carleton University case study covers how parking and security data share the same operational database. Campus-specific considerations are collected in the campus parking and security operations article.

What should a privacy review cover?

LPR collects vehicle location data, which makes privacy a design question rather than a legal formality, particularly in Canada where PIPEDA and provincial legislation set obligations for organisations collecting personal information. The practical review comes down to four things:

Retention. How long are plate reads kept, and is there a reason for that period? Reads that never lead to an enforcement or security action are the ones to justify. Retention is a configuration and governance decision the organisation makes, and OPSCOM’s platform lets a site set it to match its own policy.
Access control. Who can see plate read data? Access should follow a legitimate operational need through role-based permissions rather than being available to every staff account by default.
Purpose limitation. Data collected for parking enforcement should not be repurposed for unrelated uses without a proper basis for doing so.
Transparency. In many jurisdictions, organisations using LPR are expected to tell parkers that plate data is being collected, typically through signage at entrances and lots.

OPSCOM’s platform provides role-based access controls and configurable retention settings, which are the two controls an organisation needs to run LPR within its own privacy framework. For Canadian municipalities, campuses and healthcare sites operating under provincial privacy legislation, that is where a deployment review starts.

What does the return on LPR look like in practice?

LPR is a real investment, and the return comes from three places rather than one.

Coverage. A vehicle-mounted pass reads every plate on the route, so coverage stops being a function of how many officers are available to walk it.
Compliance. When parkers see that every vehicle is scanned on every pass, behaviour changes. The aim of time-based enforcement is turnover and availability, not ticket volume.
Administration. Reads are matched as they happen, so the reconciliation work that sits behind most manual enforcement disappears.

The clearest published outcome in OPSCOM’s client library is the Town of Perth’s 91% ticket payment rate, produced by connected enforcement and payment workflows. It is worth reading precisely: a payment rate measures the share of issued tickets that were paid, not camera performance, and it reflects one municipality’s operation. The Town of Perth case study sets out how that result was produced.

Operational data from the same reads is where longer-term value tends to show up: utilization patterns, enforcement activity by zone, and the evidence needed to argue for more capacity or better layout. That side of the platform is covered in parking data and analytics.

Explore license plate recognition in depth

Each of the following posts covers a specific aspect of LPR in more detail:

Take Command of your enforcement coverage. PL8RDR reads the plate. The platform puts the read, the record and the evidence in front of the officer who makes the decision.

Ready to see how connected LPR works in a live enforcement environment?

See it in Action – Request a Demo

License Plate Recognition (LPR): The Foundation of Modern Parking Enforcement