Skip to main content

ANPR & Smart Parking System Developer in Pakistan

SmartPark is an automatic number plate recognition and parking system I built as the sole developer on a 3-member team. It is a working prototype with a real physical barrier: a sensor triggers the Pi, the plate is recognised on-device, and the barrier opens automatically after verification.

If you are looking for an ANPR developer in Pakistan, this is a system you can watch working rather than a slide deck.

Get in touchDownload CV

How the pipeline works

The Raspberry Pi stays idle until a sensor detects a vehicle, then captures exactly three frames. YOLO locates the plate, the crop goes to a Mobile PaddleOCR variant, and the result is verified by the server. On success, GPIO signals drive a relay that operates a modified wiper-motor barrier, which auto-closes after a 10-second cool-down.

The management side

A multi-role web platform handles registrations with OTP email verification, live approval queues, camera and gate management, manual overrides for staff, and immutable entry and exit logs indexed by time, date and gate location.

What a site needs varies with camera placement, lighting and barrier hardware, so I begin with an assessment of your setup before promising accuracy figures.

Relevant case studies

Frequently asked questions

What hardware is required?

At minimum a camera, an edge computer such as a Raspberry Pi, a vehicle-presence sensor and a relay-driven barrier. Exact choices depend on the site.

Does it handle local Pakistani plate formats?

Yes. PaddleOCR was fine-tuned on custom-labelled local plates for that reason.

Can staff override the barrier?

Yes. The web platform includes staff verification overrides and full logging.

Is there a demo?

Yes. The SmartPark case study has a demo video and a link to the live dashboard.

Let's talk about your project

Email dev@abdullahaltaf.com with a few lines about the product, the timeline and the role you have in mind, or use the contact form on the homepage.