🚮 CleanVision ✨

Automated Garbage Detection and Alert System for Post Offices using YOLOv5

CleanVision is a computer vision-based system developed during the Smart India Hackathon (SIH) to automate cleanliness monitoring across post office premises. It uses a custom-trained YOLOv5 model to detect garbage in surveillance footage, trigger real-time alerts, and display actionable insights via an interactive dashboard.

The project aims to ensure hygiene, transparency, and accountability within postal divisions.

🔗 GitHub Repo: https://github.com/dopTrashTrackers/website.git

---

🧠 Key Features

• ⚙️ Garbage Detection using YOLOv5
• 🕵️ Tracks Garbage Across Frames
• 📩 Real-time Alerts via SMS & Email (Twilio)
• 📊 Dashboard to Monitor Cleanliness and Compliance
• 🏆 Ranks Post Offices Based on Performance
• 👥 Community Awareness and Engagement Section

---

🛠️ Project Workflow

graph TD
    A[CCTV Feed] --> B(Frame Capture)
    B --> C{YOLOv5 Detection}
    C -->|Garbage Found| D[Track Persistence]
    C -->|Clean| B
    D --> E{Exceeds Threshold?}
    E -->|Yes| F[Trigger Alerts]
    E -->|No| B
    F --> G[Dashboard Update]
    G --> H[Compliance Verification]

Loading

1. Image Acquisition

• Surveillance footage is captured from static CCTV cameras installed at post office premises.
• The video stream is processed into frames using OpenCV, resized, and prepared as input for the YOLOv5 model.

2. Object Detection using YOLOv5

• The YOLOv5 model processes each image frame to detect garbage items like bottle caps, paper boxes, cigarette butts, and plastic.
• Each frame is divided into grid cells where bounding box coordinates, confidence scores, and object classes are predicted.
• Non-Maximum Suppression (NMS) removes overlapping boxes, retaining only the highest-confidence detections.

3. First-Time Setup & Running the AI Model

• On first-time sign-up, users must download a Python script named Launch AI Model.py from the dashboard.
• When executed, the script:
  • Captures or accepts an image input
  • Runs the YOLOv5 model locally
  • Returns detection results

🛡️ This design ensures model execution is performed locally, avoiding real-time hosting, enhancing privacy and reducing infrastructure costs.

4. Garbage Tracking & Alert System

• Garbage is tracked across frames using bounding boxes and size measurement.
• If garbage persists and exceeds a predefined threshold:
  • An automated alert is triggered.
  • Twilio SMS and Email are sent to divisional officers and cleaning staff.
  • Includes a proof-submission link for uploading a cleaned image post-action.
  • Dashboard updates in real-time with compliance status.

5. Interactive Dashboard

Built using React.js, the dashboard offers a dynamic, real-time view of activity.

📈 Visual Analytics (Powered by Recharts)

• Bar Chart – Frequency of different garbage types.
• Line Chart – Trends of garbage detection over time.
• Pie Chart – Compliance scores per post office.
• Tabular View – Timestamped list of detection events.

🏢 Best vs Worst Performing Post Offices

• An interactive map highlights top and bottom performers based on compliance scores.

🌍 Community Initiatives

• Showcases cleanliness drives, awareness programs, and encourages community involvement.

📊 Real-time Ranking System

• Ranks each post office based on:
  • ⏱️ Response time to alerts
  • 🧹 Participation in community programs
  • 🔄 Frequency of garbage detections

---

📦 Tech Stack

Layer	Tools
Frontend	React.js, HTML, CSS, Recharts
Backend	Flask, Firebase (Authentication & Storage)
Notifications	Twilio API (SMS & Email)
Mapping & Charts	React Leaflet (Map), Recharts

---

🚀 Getting Started

1. Clone the Repository

git clone https://github.com/dopTrashTrackers/website.git
cd website

🚀 Getting Started

2. Install Frontend Dependencies

Make sure you have Node.js and npm installed. Then, install all frontend packages and start the development server:

npm install      # Installs all frontend dependencies
npm run dev      # Starts the React development server (Vite)

3. First-Time User: Run the AI Model Locally

CleanVision uses a YOLOv5 model to detect garbage from captured images. For privacy and efficiency, the model runs locally on your machine.

🔽 Step A: Download the Inference Script

Download the Python script named: Launch AI Model

🧪 Step B: Install Python & Model Dependencies

Make sure you have the following installed on your system:

• Python 3.7+
• YOLOv5 repository and its dependencies
• OpenCV
• PyTorch
• Other dependencies: numpy, Pillow, matplotlib

You can install everything using:

pip install -r requirements.txt

If requirements.txt is not available, install dependencies manually:

pip install torch torchvision torchaudio
pip install opencv-python numpy matplotlib pillow

Step C: Run the Script Execute the downloaded Python script in your terminal:

python "Launch AI Model.py"

📬 Contributing

We welcome contributions to improve CleanVision! Follow the steps below to get started:

🛠️ Step 1: Fork the Repository

Click the Fork button at the top-right of this page to create your own copy of the repository.

🌱 Step 2: Create a New Branch

git checkout -b feature/your-feature-name

✍️Step 3: Make Your Changes & Commit

Edit the codebase as needed and commit your changes:

git commit -m "Added new feature"

🚀 Step 4: Push to Your Fork

git push origin feature/your-feature-name

📥 Step 5: Create a Pull Request

Open a Pull Request from your forked repository back to the main branch of the original project.

🏁 Acknowledgments

🏆 Finalist – Smart India Hackathon 2024

🧹 Inspired by the Swachh Bharat Mission

👨‍💻 Team Members

🌐 Web & Dashboard Development	🤖 AI/ML Model Development
Souvik Maity	Nidhi Satyapriya
Abhijeet Awasthi	Rahul Kumar
	Shraddha Sahu
	Shiekh Mahammad Arzu