AI-Powered Vehicle Intelligence

ANPR:
Automatic
Number Plate
Recognition
+ Telegram/Whatsapp Messages

How AI, Computer Vision & OCR Transform Vehicle Identification — from toll booths to smart cities and beyond.

Explore How AI-enabled Security See Applications
1000 visitors
98%
Recognition accuracy
<1s
Processing time
24/7
All-weather operation
ANPR system detecting a Malaysian licence plate
Scroll
Understanding the Technology

What is ANPR?

Automatic Number Plate Recognition (ANPR) combines high-resolution cameras, optical character recognition (OCR), and machine learning to instantly read and process vehicle licence plates — with no human intervention required.

Used globally across industries, ANPR enables seamless automation of processes that once required manual checks: cashless toll collection, smart parking, warehouse loading bays, vehicle entry & exit logging, and community safety management.

Traffic Management Warehouse Loading Security Community Safety Vehicle Entry / Exit
See how it works
ANPR camera system at a gate
Plate Detected
WA 1234 B
AI
Powered
The Pipeline

How ANPR Works

From camera capture to gate action — a 5-step AI pipeline that completes in under one second.

01
📸

Image Capture

High-res IR camera captures the vehicle at trigger point or via continuous video stream.

02
🔍

Plate Detection

YOLOv8 deep-learning model locates and crops the plate region from the full frame.

03
🔠

OCR Recognition

EasyOCR extracts characters with regex-scored validation across multiple passes.

04
🗃️

Database Match

Plate cross-referenced against SQLite whitelist of authorised vehicles.

05
⚡

Action Triggered

Gate opens for whitelisted plates. Unknown vehicles trigger Telegram / WhatsApp alert. All events logged.

< 1 second
Total end-to-end processing time — image capture to action
Capture → OCR: ~200 ms DB lookup: ~50 ms MQTT relay: ~30 ms
ANPR systems in various real-world applications
Use Cases

Key Applications

ANPR technology powers critical infrastructure across every industry segment.

🛣️

Toll Collection

Cashless, barrier-free toll automation eliminates queuing and manual payment. Vehicles are billed automatically via linked accounts.

Highway & expressway
🅿️

Smart Parking

Automated entry/exit and duration-based billing. Vehicles enter and exit without stopping — the system handles everything in the background.

Malls & commercial zones
🚔

Law Enforcement

Instant cross-reference against stolen vehicle databases, expired registrations, and wanted persons. Real-time alerts to patrol officers.

Police & JPJ
🏙️

Smart City Traffic

Real-time flow monitoring, congestion analysis, and origin-destination studies. Data feeds into adaptive traffic signal systems.

Urban planning
🏢

Access Control

Whitelist-based vehicle access for gated communities, corporate campuses, and restricted zones. Integrates with existing boom gates.

Residential & corporate
🌐

Border Security

Cross-border vehicle verification against international databases. Handles multi-format plates from different countries automatically.

Customs & immigration
Why Deploy

Transformative Benefits

⚡

Whitelist Database

Authorised vehicles granted instant access. Non-whitelisted plates trigger automatic alerts — zero manual verification needed.

🎯

Auto Messaging

Instant Telegram or WhatsApp notifications with plate photo, timestamp, and gate location sent automatically when unknown vehicles are detected.

🌙

24/7 All-Weather

IR illumination and adaptive thresholding ensure reliable detection in rain, night-time, glare, and adverse conditions — with full database recording.

Cost Reduction

Eliminates manual toll and parking staff. One system handles thousands of vehicles per day, 24/7, with minimal maintenance overhead.

Scalability

Deploy from a single gate to a city-wide multi-camera network. Cloud architecture scales horizontally with no redesign required.

Integration

Works with IoT sensors, cloud platforms, and existing CCTV infrastructure via MQTT and REST API output streams out of the box.

RGB LED Display

ESP32-S3 driven 192×32 outdoor P5 panel shows the detected plate in real-time via MQTT — instant visual confirmation at the gate.

Deployment Types

Fixed vs Mobile Systems

Choose the deployment model that best fits your infrastructure and operational requirements.

Fixed Systems

Permanent Installation
Fixed ANPR camera at toll booth
  • Permanent infrastructure at toll booths, parking entrances, and border crossings
  • Optimised illumination and angle for consistent, high-accuracy reads
  • 24/7 continuous monitoring with local edge processing
  • Integrated boom gate / barrier relay control via MQTT
Toll Booths Parking Borders Gated Communities

Mobile Systems

Vehicle-mounted Units
Mobile ANPR on police patrol vehicle
  • Mounted on police and enforcement patrol vehicles for dynamic coverage
  • Scans plates of passing vehicles while the patrol car is in motion
  • Flexible deployment at events, roadblocks, or high-crime areas on demand
  • Connected to national databases via 4G/LTE for real-time officer alerts
Police JPJ Enforcement Event Security Roadblocks
🇲🇾 Local Context

ANPR for Malaysian Roads

Tailored to Malaysian plate formats, JPJ database standards, and the unique traffic patterns of our roads.

Malaysian Plate Formats Supported

WA 1234 B

Kuala Lumpur — Standard

BK 123

Sabah — Short format

JHB 9999

Johor — State prefix

A 1234

Single-letter prefix

Regex: [A-Z]{'{1,3}'}\\d{'{1,4}'}[A-Z]{'{0,2}'}

Handles standard, short, and all state-prefix formats

Malaysian Use Cases

Touch 'n Go / MyEG Style Toll

Multi-lane free-flow toll collection with automatic account deduction. Plate acts as the vehicle identifier, removing the need for RFID tags.

JPJ Database Integration

Cross-reference plates against Jabatan Pengangkutan Jalan (JPJ) records for roadworthiness, ownership, and registration status verification.

Telegram / WhatsApp Alerts

Security guards and condo management receive instant notifications with plate photo when an unknown vehicle enters the premises.

Gated & Guarded Communities

Replace manual logbooks with automated ANPR — residents pass through automatically; visitors trigger guard notification via Telegram.

Malaysian highway ANPR deployment

Replace with real Malaysian toll / condominium gate photo

State Plate Prefixes — All Supported

W — KL B — Selangor J — Johor P — Penang N — N. Sembilan A — Perak BK — Sabah QA — Sarawak VD — Putrajaya
Under the Hood

The Modern ANPR Tech Stack

Full-Stack Pipeline Architecture
YOLOv8 Detection OpenCV Preprocessing EasyOCR / Tesseract SQLite / PostgreSQL MQTT / REST API

The pipeline begins with YOLOv8 nano detecting vehicle bounding boxes, followed by OpenCV preprocessing (grayscale, histogram equalisation, morphological operations) applied in multiple passes to the cropped plate region. EasyOCR runs character recognition with regex-scored candidate validation. Confirmed plates are persisted to SQLite and published over MQTT to edge devices — including ESP32-S3 RGB LED matrix displays and Telegram bots.

AI / ML Layer

  • YOLOv8n — vehicle detection
  • EasyOCR — multi-language OCR
  • Regex scoring — plate validation
  • Multi-pass preprocessing

Data / Backend

  • SQLite — whitelist & log storage
  • CSV — parking log export
  • JSON config — easy customisation
  • PostgreSQL — enterprise scale

Connectivity

  • MQTT — IoT event publishing
  • Telegram Bot API — alerts
  • WhatsApp notifications
  • REST API — system integration

Camera / Vision

  • IP camera via RTSP / HTTP
  • USB webcam (V4L2)
  • Android IP Webcam app
  • Docker + X11 display support

Edge Hardware

  • ESP32-S3 N8R16 — MQTT subscriber
  • P5 RGB LED 192×32 display
  • Raspberry Pi — local broker
  • Mosquitto MQTT broker

Software Stack

  • Python 3 — main runtime
  • OpenCV (system pkg) — imaging
  • paho-mqtt — MQTT client
  • Docker / WSL2 deployment

System Architecture — Data Flow

IP Camera
Python ANPR
SQLite DB
MQTT Broker
ESP32-S3
P5 LED Matrix
Telegram / WhatsApp alerts dispatched in parallel for unknown plates
About Us

Palvision Ledtronics

This system was developed independently and locally by a coding teacher, driven by a passion for accessible AI and smart automation solutions for Malaysian communities.

New AI features will be continuously updated and tailored according to customer needs — ensuring the system evolves alongside the latest advancements in computer vision and plate recognition technology.

We welcome constructive feedback from users, communities, and partners to help us build a better, smarter, and more reliable ANPR experience.

Penang Georgetown landmark
Georgetown, Penang — Where It All Began