AI-Powered Number Plate Recognition System for Indian Vehicles – with State Detection & OCR Magic!
VEHICLE-NUMBER-PLATE-DETECTION is an AI-powered Flask web application that:

✅ Detects vehicle number plates from images using TensorFlow's SSD model
🧠 Extracts text using EasyOCR
📍 Identifies the Indian state based on the license plate
🌐 Provides a web interface ready to deploy on Pivotal Cloud Foundry
All powered by machine learning & computer vision — like magic, but real!
| Component | Technology Used |
|---|---|
| Object Detection | TensorFlow (SSD MobileNet) |
| OCR | EasyOCR |
| Web Framework | Flask |
| Image Processing | OpenCV |
| Deployment Ready | Pivotal Cloud Foundry |
| Language | Python |
| AI Feature | Description |
|---|---|
| 🎯 Object Detection | Detects number plates using SSD MobileNet |
| ✍️ OCR | Extracts alphanumeric text using deep learning (EasyOCR) |
| 🗺️ State Mapping | Maps license prefix to Indian states (e.g., MH → Maharashtra) |
| 🖥️ Flask UI | Clean dashboard to upload and scan vehicle images |
Here's what a real-time dashboard could look like in your implementation:
- Minimum Image Size: 800x600 px (for optimal accuracy)
- Supported Formats:
.PNG,.JPG,.JPEG - Clear view of the front/rear number plate
git clone https://github.com/yourusername/VEHICLE-NUMBER-PLATE-DETECTION.git cd VEHICLE-NUMBER-PLATE-DETECTION
python app.py Now open your browser and go to: http://localhost:5000
💻 Demo Flow 1️⃣ Upload Vehicle Image ↓ 2️⃣ Plate Detected via TensorFlow SSD ↓ 3️⃣ OCR reads the Number Plate ↓ 4️⃣ State Decoded from Prefix ↓ 5️⃣ Results shown on Dashboard!
✅ Ready for Pivotal Cloud Foundry, Heroku, Render, or even Docker. Just add your Procfile, manifest.yml, and you're good to go.
🎥 Real-time webcam feed support
🧑
🚗 Vehicle make/model prediction
🛰️ GPS integration for geo-tagging
📱 Mobile-friendly PWA version
Input Image Plate Text State car1.jpg MH12AB1234 Maharashtra vehicle_2.jpg DL05XZ3342 Delhi


