People Detector 👥

A Flask-based web application that detects people in camera surveillance images using OpenCV's HOG (Histogram of Oriented Gradients) descriptor.

Features

• Upload images through a clean web interface.
• Automatic people detection using HOG + SVM
• Visual results with bounding boxes around detected people
• Responsive and modern UI

Technologies Used

• Flask: Web framework
• OpenCV: Computer vision and people detection
• Gunicorn: Production WSGI server
• HTML/CSS: Frontend interface

Web Deployment

Click link and insert .png or .jpg image file of a surveillance footage of a crowd in traffic. https://peopledetector-1.onrender.com

Prerequisites

• Python 3.8 or higher
• pip

Project Structure

PeopleDetector/
├── app.py                 # Main Flask application
├── detect.py             # People detection logic
├── requirements.txt      # Python dependencies
├── render.yaml          # Render deployment config
├── .gitignore           # Git ignore rules
├── static/
│   ├── style.css        # CSS styling
│   ├── uploads/         # Uploaded images (not tracked)
│   └── results/         # Detection results (not tracked)
└── templates/
    └── index.html       # Main HTML template

How It Works

1. User uploads an image through the web interface
2. Image is saved to the static/uploads/ directory
3. OpenCV's HOG descriptor detects people in the image
4. Bounding boxes are drawn around detected people
5. Result image is saved to static/results/ directory
6. Result is displayed to the user

Notes

• The app uses opencv-python-headless for Render deployment (no GUI dependencies)
• Upload and result directories are created automatically if they don't exist
• Old uploaded/result images are not automatically cleaned up

License

MIT License - feel free to use this project for learning and development!