Optical Plate Recognition (ALPR Technology)
Automatic license plate recognition (ALPR) uses cameras and software to capture a vehicle’s plate image and convert its characters into searchable data. A dependable, lawful system starts with a defined purpose, a site survey, and tests at real distances, speeds, and lighting conditions. Camera resolution alone cannot guarantee accurate reads; measure plate detail and verify results on site.
If you are planning a system, or trying to understand one already in use, the key question is not simply how many megapixels a camera has. It is whether the system captures a readable plate under the conditions it will face, and whether it handles the resulting data lawfully and securely.
I approach ALPR as a complete video system: camera, lens, lighting, processing, network, and rules for access and retention. A useful plan defines what the system must do, where it may capture vehicles, and how performance will be checked before use.
Planning, Lawful Purpose, and Site Survey
A site survey records the purpose and conditions that shape the system design. Before choosing equipment, document where vehicles travel, which plates the system must read, and what rules govern access and retention. This prevents a technical design from collecting more information than its approved purpose requires.
Start by defining the capture zone: the specific lane or area where plates should be read. Record vehicle direction, typical and maximum speed, capture distance, likely plate jurisdictions, lighting, and any limits on mounting or cabling. Note who may access plate records, why they need access, and how long records may be kept under applicable law and policy.
Privacy and system performance are linked. A camera aimed too broadly may collect more than needed, while a poorly chosen capture zone may miss plates. Limit the field of view to the approved area where practical, and document the reason for any wider view. Set retention limits before operation, and check that the system can apply them.
IEC 62676-4:2014 provides guidance for planning video-surveillance systems. Its DORI values describe general levels of image detail: detection at 25 pixels per metre, observation at 62.5, recognition at 125, and identification at 250. These are planning references, not ALPR accuracy thresholds. They do not replace measuring plate detail or testing the installed system.
Survey checklist
- Define the lawful purpose, capture zone, and traffic direction.
- Measure near and far capture distances and record vehicle speeds.
- Identify lighting changes, glare, headlights, and likely obstructions.
- Confirm mounting, service access, and safe cable routes.
- Record network, power, privacy, access, and retention requirements.
The first decision is not which camera to buy. It is what the system is authorized and expected to capture.
Field of View, Mounting, and Access Clearances
Field of view is the area a camera can see at a given lens and distance. Mounting angle, height, and distance affect how large and clear a plate appears in the image. A design should confirm plate detail at both the closest and farthest capture points, not just show that the lane is visible.
Ask the supplier for the minimum number of pixels across a plate required by its software, then verify that figure at the farthest capture point. There is no single pixel-width rule that applies to every camera, plate design, jurisdiction, or recognition algorithm. Image quality, motion, exposure, and software also affect results.
Check the view from the camera’s actual position. A plate can be partly hidden by another vehicle, a trailer, a bumper, dirt, or an unusual mounting position. A steep viewing angle can distort characters. Glare, reflective surfaces, and direct headlight exposure can also reduce usable detail. Confirm sightlines during the hours and conditions the system is meant to cover.
| Design check | What to measure or verify | Why it matters |
|---|---|---|
| Plate coverage | Pixels across a plate at the farthest point | Confirms detail against the vendor’s stated requirement |
| Capture distance | Near and far limits in metres | Prevents testing only at an easy distance |
| Vehicle speed | Typical and maximum speed | Motion can blur characters |
| Camera angle | Horizontal and vertical angle to the plate | Affects visible character shape |
| Sightline | Likely vehicle and lane obstructions | Helps identify missed or blocked plates |
IEC DORI values can help frame general video detail, but plate pixels should be checked directly. A high-resolution camera may still devote too few pixels to a distant plate if its lens covers a very wide area. Conversely, a narrower view may provide more plate detail but cover less of the site.
Plan service access as well as the view. The camera must be reachable for focus checks, cleaning, and maintenance without placing staff or the public at risk. Keep the installation within site rules and equipment limits. The next step is to select and configure equipment around measured conditions.
Camera, Power, Network, and Configuration
An ALPR camera produces images for plate-reading software, either within the camera or on a separate system. Selection should match the measured capture zone and the vendor’s requirements. Confirm model-specific features, power needs, and interoperability rather than relying on broad claims such as “AI-ready” or “works with any recorder.”
Choose lens and resolution using the plate-size calculation and on-site view, not resolution alone. Test focus at the working distance. Exposure, shutter, and illumination must suit the vehicle speed and ambient light. A shutter setting such as 1/1000 second can be a starting point for testing, but it is not a universal requirement. Final settings depend on the camera, speed, distance, and light.
Infrared (IR) light can support night capture, but reflective plates can become overexposed under IR. Characters may wash out even when the surrounding scene looks well lit. Test the plate itself at the operating distance and speed. Adjust camera angle, exposure, or illumination, or use a suitable polarized or filtered setup if the equipment supports it.
| Component | Verify before installation | Practical check |
|---|---|---|
| Camera and lens | Supported plate detail and operating range | Check plate pixels at near and far limits |
| IR or white light | Supported modes and safe operating limits | Test plates at night and under headlights |
| Power | Listed camera demand and power-source rating | Allow for cable loss and device needs |
| Network | Bandwidth, secure transport, and time sync | Confirm stable video and correct timestamps |
| Recorder or software | Exact supported models and features | Test plate data, events, and playback |
Power over Ethernet (PoE) sends power and data over one Ethernet cable. IEEE 802.3af Type 1 allows up to 15.4 watts at the power-sourcing equipment; IEEE 802.3at Type 2 allows up to 30 watts there. Those figures are not guaranteed at the camera. Check the camera’s listed demand and account for cable-run losses.
ONVIF Profile T covers supported IP-video functions, but it does not mean every camera, recorder, or ALPR software feature will work together. Check the exact models and their conformance details. In particular, confirm whether the system supports the plate metadata and events you need, not only basic video.
Use secure transport, unique user accounts, and access roles that match approved duties. Keep system time synchronized so image timestamps and records can be compared. Do not expose feeds or credentials to public networks. Document the approved network path and who can change settings or retrieve records.
Acceptance Testing, Security, and Retention Verification
Acceptance testing checks whether the installed system works under its intended conditions. A successful live view is not enough: the team must assess plate capture and reading across representative speeds, distances, lighting, and plate cases. Record the test method, results, failures, and changes so performance can be reviewed later.
Test day and night, near and far capture points, and the expected range of vehicle speeds. Include wet weather if it is part of the operating environment, as well as headlights and likely glare. Test representative plate types and jurisdictions where law and site purpose permit. Record how many plates were presented, captured, and read correctly; define the counting method so results can be compared.
Do not treat a vendor’s headline accuracy figure as a promise for your site. Results depend on the camera, setup, environment, plates, and test method. Nor do DORI values or camera resolution prove ALPR performance. If the system misses plates or misreads characters, check alignment, focus, exposure, light, and the software’s configuration before accepting it.
Illustrative commissioning scenarios, not reported case studies
- Daylight lane: A test vehicle passes at near and far points. If far plates have too little detail, review the lens and field of view, then repeat the test.
- Night approach: Headlights and IR are tested at the operating distance. If reflective characters wash out, adjust exposure, angle, or supported lighting options, then retest.
- Network or recorder change: Video remains visible after a device change, but plate events do not appear. Verify exact model support and metadata features before treating the integration as complete.
After acceptance, confirm that user roles, access logs, and retention settings match the approved policy and applicable law. Check that records are removed on schedule and that only authorized people can retrieve or share them. Keep a record of approved access and any changes to the system.
Commissioning record
- Camera, lens, software, and recorder model details.
- Capture distances, speeds, lighting, and test conditions.
- Plate counts, reads, missed captures, and known failure modes.
- Final exposure, focus, illumination, and time settings.
- Approved users, network path, and retention rules.
A documented test provides a baseline. Repeat checks after major changes to camera position, software, lighting, or the lane.
Conclusion and FAQ
A reliable plate-reading system is planned around a lawful purpose and verified under real operating conditions. Measure plate detail at the farthest point, test exposure and lighting at actual speeds, and confirm that power, network, software, access, and retention settings work as intended. Keep the test record so later changes can be checked against the same baseline.
What does ALPR do?
ALPR captures an image of a vehicle plate and uses software to convert visible characters into data. Depending on the system, it may associate that data with an image, timestamp, or camera location.
Does a higher-resolution camera guarantee accurate plate reads?
No. Resolution alone does not show how many pixels cover a plate or whether motion, glare, angle, or exposure will obscure its characters. Test the installed camera at the actual capture distances and speeds.
How many pixels across a plate are needed?
There is no universal number for all systems. Ask the software supplier for its minimum plate-pixel requirement and verify that detail at the farthest capture point, using the intended camera, lens, and capture conditions.
What do DORI values mean for plate recognition?
DORI values are general video-planning references for detection, observation, recognition, and identification. IEC 62676-4:2014 lists 25, 62.5, 125, and 250 pixels per metre for those stages. They are not ALPR accuracy guarantees.
Why can a plate look washed out at night?
Many plates reflect light strongly. IR illumination can overexpose the plate and hide characters, even when the rest of the scene looks clear. Check exposure on plates at operating distance and speed, then adjust supported settings.
Is a 1/1000-second shutter speed always required?
No. It can be a starting point for testing, but the right exposure depends on speed, distance, light, camera capability, and the desired image. Test at the site rather than treating one setting as universal.
Does ONVIF Profile T ensure full ALPR compatibility?
No. Profile T covers supported IP-video functions, but specific devices may not support every ALPR event or metadata feature. Check the exact camera, recorder, and software models and test the required data flow.
How much power does a PoE camera receive?
IEEE 802.3af Type 1 allows up to 15.4 watts at the power source, and 802.3at Type 2 up to 30 watts there. The amount available at the camera can be lower. Check device demand and cable losses.
How should an ALPR system be tested before use?
Test representative distances, speeds, day and night conditions, headlights, and relevant weather and plate cases. Record total plates, captures, correct reads, and failure modes using a consistent method.
How long should plate records be retained?
Retention depends on the authorized purpose and applicable local law or policy. Set a clear limit, restrict access to approved users, and verify that records are deleted on schedule.
(This article was written by one of our staff writers, Steven Langford. Visit our Meet the Team page.)
