SquintPRO · Vehicles
ANPR system: number plate recognition and vehicle analytics with SquintPRO
An ANPR (automatic number plate recognition) system reads vehicle number plates from camera images and turns them into text. That text can then be checked against a list to open a gate, record a parking stay or log a delivery. Reliable plate reading depends on the camera as much as the software: angle, plate size in the image, shutter speed and night lighting decide most of the result. This page sets out what to require from an ANPR system in Saudi Arabia. It also covers what SquintPRO, the video analytics platform Interkey builds, does with vehicles. SquintPRO reads number plates on cameras positioned for plate reading. On the cameras you already have, it detects vehicles, raises restricted-area alerts, tracks vehicles across cameras and searches footage.
Published 5 min read
In short
- ANPR reads plates. It needs cameras placed and set up for plates, usually one lane per camera at a gate or barrier.
- Saudi plates carry two scripts. They show the registration in Arabic and in Latin characters, so ask how a system reads and checks both.
- Vehicle analytics answer different questions. Where did the vehicle go, did it enter a restricted area, and when did a similar vehicle pass?
- SquintPRO: reads plates at gates on cameras positioned for plate reading. On existing CCTV it detects vehicles and people, alerts when a vehicle is where vehicles should not be, tracks vehicles across cameras on a site map, and searches footage by description or image.
- No face recognition. SquintPRO does not identify people.
How an ANPR system works
- Capture. A camera takes an image of the vehicle as it approaches a gate, barrier or lane.
- Find the plate. Software locates the plate within the image.
- Read the characters. The plate is converted into text, character by character.
- Match. The text is checked against lists: permitted vehicles, expected deliveries, or vehicles already inside.
- Act and record. The system opens a barrier, raises an alert or writes an entry and exit record, with the image as evidence.
Most problems happen at the first and third steps. The camera did not capture a readable plate, or one character was misread.
Saudi plates: what the system must read
Saudi plates show the registration in both Arabic and Latin letters and numerals. When you assess a system, ask:
- Does it read both scripts, and does it check one against the other?
- How does it record the plate: in Arabic, in Latin characters, or both?
- How does it handle a dirty, damaged or partly hidden plate? Does it flag the read as uncertain, or guess?
- How does it handle plates from other GCC countries, if they enter your site?
What plate reading needs from the camera
- Angle. The camera should face the plate as directly as possible. Steep angles distort the characters.
- Plate size in the image. The plate must be large enough for each character to be distinct. A wide overview camera rarely achieves this.
- Shutter speed. Moving vehicles blur at slow shutter speeds, especially at night.
- Night lighting. Infrared illumination is common for plate cameras, because headlights and darkness defeat ordinary views.
- One lane per camera where possible, so two vehicles do not overlap.
This is why ANPR is usually installed at specific points, such as gates, barriers and car-park entries. It is not added to every overview camera on a site.
Where ANPR and vehicle analytics are used
| Need | What answers it | What to require |
|---|---|---|
| Open a barrier for permitted vehicles | ANPR at the gate | Read accuracy measured at your gate, day and night, and a fallback when a read fails |
| Log every vehicle in and out of a yard | ANPR at entry and exit points | Records that link the image, time and lane, and a retention period you set |
| Know when a vehicle enters a restricted area | Vehicle detection with zone rules | Alerts with video evidence, routed to your control room |
| Reconstruct a vehicle's route inside a site | Cross-camera vehicle tracking | Camera positions on a site map, and a person who verifies each step |
| Find a vehicle in recorded footage | Search by description or image | Search across the cameras you choose, within a time window |
What SquintPRO does with vehicles today
SquintPRO works on the views the camera survey qualifies. Plate reading needs cameras positioned for plates; the other functions run on the CCTV you already have:
- Reads number plates at gates and entry points, on cameras positioned for plate reading, and logs each vehicle's entry and exit with the plate, time and image.
- Detects vehicles and people in each camera view.
- Alerts when a vehicle is where vehicles should not be, such as a pedestrian area or a restricted zone, during the hours you set. Each alert carries video evidence and goes to your existing video management, control-room or incident system.
- Tracks a vehicle across cameras, with each camera's position on a map of the site (GIS), so investigators can see the route.
- Searches recorded footage by description or by a reference image, for example "a white pickup in the last three hours".
- Lets you define new events by text or a reference image, without retraining. Test each new definition on your own footage.
- Logs every event with timestamp, camera, zone and event type.
Read more
SquintPRO runs on-premise, so footage stays on your site, or in the cloud with 3, 6 or 12 months' retention. It is priced per use case. Each quote is built from the site, the number of cameras and the use cases in scope.
Questions to ask any ANPR supplier
- What read rate do you measure, on which plates, and under what conditions? Can we measure it at our own gate?
- Does the system read both scripts on Saudi plates, and check one against the other?
- Which cameras does it need, and can any of our existing cameras be used?
- How does it connect to our barrier, access-control and video management systems?
- Where are plate images and records stored, who can see them, and for how long?
- What happens when a read fails: manual entry, intercom, or an alert to the guard?
Data handling
Plate records describe where and when a vehicle was seen. Treat them with the same care as other records about people. Decide the purpose, who can search the records, how long they are kept and how drivers are informed, before go-live. With on-premise deployment, SquintPRO processing runs on your infrastructure and footage does not leave the site.
Questions buyers ask
Can SquintPRO read number plates?
Yes, on cameras positioned for plate reading at gates and entry points. Measure the read rate at your own gate, day and night, during the pilot.
Do we need dedicated cameras for ANPR?
In most cases, yes, at the points where plates are read. Plate reading needs a direct angle, a large enough plate image, a fast shutter and night lighting, which overview cameras rarely provide.
Can SquintPRO track a vehicle across our site?
Yes. SquintPRO tracks vehicles as well as people across cameras, and shows each camera's position on a map of the site. A person verifies each step.
Can we find a vehicle in recorded footage without knowing the camera?
Yes. Search by description or with a reference image across the cameras SquintPRO analyses, within a time window you choose.
Does SquintPRO identify drivers?
No. SquintPRO does not use face recognition.
Where should we start?
With the question you need answered. Gate access and entry logs point to ANPR at fixed points. Restricted-area alerts, routes and investigations point to vehicle analytics on the cameras you already have. Many sites need both, and SquintPRO covers both.
Next
Related pages
Next step
Discuss vehicle analytics for your site
Tell us about the site, its gates and vehicle areas, and roughly how many cameras cover them. We will tell you where plate reading needs dedicated cameras, which needs your existing cameras can meet, and what a pilot could measure.
Or directly
+966-11-2180999 [email protected]Tawuniya Towers, North Tower, 7th Floor, King Fahad Highway, Olaya, P.O. Box 56835, Riyadh 11564, Saudi Arabia