🍽️ Plate Monitor – YOLOv8 + Flask Demo

This project is an upgraded version of a Smart India Hackathon prototype.
It uses YOLOv8 (Ultralytics) for object detection combined with a Flask backend and a simple frontend
to analyze plate images, estimate food coverage, and assign a "fine" based on leftovers.

---

🚀 Features

• Upload an image of a plate via web UI
• YOLOv8-powered object detection for food & plates
• Calculates coverage % (food area / plate area)
• Assigns plate status (empty, half-full, etc.) with fines 💸
• Results displayed with processed image & detected items list
• JSON API available (for programmatic access)

---

📂 Project Structure

├── app.py                 # Flask backend (routes & upload handling)
├── plate_detection.py     # YOLOv8 + CV logic for analyzing plates
├── requirements.txt       # Dependencies
├── templates/
│   └── index.html         # Frontend template (Jinja2)
├── static/
│   ├── style.css          # Styling for UI
│   └── outputs/           # Processed images (auto-created)
├── uploads/               # Uploaded images (auto-created)
└── README.md              # Documentation

---

⚙️ Installation

1️⃣ Clone Repository

git clone https://github.com/ohnogaurav/plate-monitor.git
cd plate-monitor

2️⃣ Create Virtual Environment

python -m venv venv
source venv/bin/activate   # On Linux/Mac
venv\Scripts\activate    # On Windows

3️⃣ Install Requirements

pip install -r requirements.txt

---

▶️ Usage

Run Flask App

python app.py

By default, app runs at: http://127.0.0.1:5000/

Upload Plate Image

• Go to web UI in browser
• Upload .jpg / .jpeg / .png file
• View processed result with detected items + coverage

API Usage (JSON)

Send POST request with image file:

curl -X POST -F [email protected] http://127.0.0.1:5000/upload -H "Accept: application/json"

Response example:

{
  "status": "Plate is half full",
  "fine": 35,
  "coverage_percent": 47.5,
  "detected_items": [
    {"label": "plate", "conf": 0.92},
    {"label": "spoon", "conf": 0.88}
  ],
  "output_image_url": "/static/outputs/processed_1692628234_test.jpg"
}

---

📊 Tech Stack

• Backend: Flask (Python)
• Computer Vision: YOLOv8 (Ultralytics), OpenCV
• Frontend: HTML, CSS (Jinja2 templating)
• Others: Werkzeug for uploads, REST-style JSON API

---

👨‍💻 Author

Project developed & upgraded by Gaurav Kumar
Originally built during Smart India Hackathon