🔍 Real-Time License Plate Detection & Recognition 🚗📷

A smart, real-time web-based system built using YOLOv8 + PaddleOCR for detecting and recognizing Indian vehicle number plates from images. Supports automatic preprocessing, validation, and error correction of license plate formats.

---

⚙️ Tech Stack

• 🔧 Flask – Lightweight web framework for the frontend/backend.
• 🧠 YOLOv8 – Custom-trained object detection model (best2.pt) for number plate detection.
• 📝 PaddleOCR – Extracts license numbers from the detected plates.
• 📸 OpenCV – Image preprocessing for clarity and thresholding.
• 🔡 Regex & DiffLib – Validates and autocorrects plate text formats.

---

🚀 Features

• 📸 Upload an image/video or start webcam to detect number plates in real-time.
• 🧽 Auto-enhancement of plates before OCR.
• 🧾 Smart validation & correction of state codes (e.g., MH, DL, KA).
• 📤 Accepts .jpg, .jpeg, .png image files.
• 🌐 Auto-detects and runs on an available local port.

---

📁 Project Structure

├── Main-Final.py        # Main Flask app using YOLO + OCR
├── best2.pt             # Custom-trained YOLOv8 model (keep in same directory)
├── templates/
│   └── index.html       # Web UI template
├── uploads/             # Temporary uploaded files
└── README.md            # Project description file

---

🔧 Setup & Run

1. 🔽 Clone the repo

   git clone https://github.com/kenreddy23/real-time-license-plate-detection-and-recognition.git
   cd real-time-license-plate-detection-and-recognition

2. 📦 Install dependencies

   pip install -r requirements.txt

3. 🧠 Make sure best2.pt is in the same folder as Main-Final.py.

4. 🚀 Run the app

   python Main-Final.py

5. 🌐 Open the local server URL shown in the terminal to test it.

---

✨ Sample Output

---

🛠 Built with Python, YOLO, PaddleOCR... and a lot of coffee ☕🐍