Usagi-AI-

Helmet & License Plate Detection 🚦🪖

This project combines YOLOv8 object detection with OCR (EasyOCR / Tesseract) to detect:

• Whether a rider is wearing a helmet or not wearing a helmet.
• Recognize and extract the vehicle license plate number from images.

The project runs seamlessly on Google Colab.

---

📌 Features

✅ Helmet detection (Helmet / No-Helmet classification)
✅ License plate detection using YOLO bounding boxes
✅ License plate number recognition using OCR (Tesseract / EasyOCR)
✅ Works on both images and real-time video feeds (with minor tweaks)
✅ Output shows bounding boxes with labels & recognized plate number

---

🛠️ Tech Stack

• 🐍 Python 3
• 🚀 YOLOv8 (Ultralytics)
• 🎥 OpenCV
• 🔎 EasyOCR / Tesseract OCR
• ☁️ Google Colab (for running without local setup)

---

📂 Project Structure

Helmet-Plate-Detection/
│
├──weights
📦 best.pt              # YOLO trained model weights (License Plate)
🪖 helmet_best.pt        # YOLO helmet detection weights
├──app..py                #python file
├── 🖼️ demo_image.jpg        # Sample input image
├── 📒 app.ipynb           # Google Colab notebook
├── 📜 requirements.txt     # Python dependencies
└── 📘 README.md            # Project documentation

2️⃣ Install Dependencies

Make sure you are inside the project folder, then install dependencies from requirements.txt:

cd Helmet-Plate-Detection
pip install -r requirements.txt

3️⃣ Run on Google Colab

	•	Upload best.pt (license plate detection weights)
	•	Upload helmet_best.pt (helmet detection weights)
	•	Open usagi.ipynb in Colab and run the cells