A real-time person detection system that uses a trained RT-DETR model through Roboflow Workflows and provides a simple web interface for monitoring the camera feed.
The project captures video from a laptop camera and sends the frames through the Roboflow inference workflow. The processed video is then displayed on a Flask-based web interface along with the detected person count and FPS.
- Real-time person detection
- RT-DETR based object detection
- Roboflow Workflow integration
- Flask web interface
- Live camera feed
- Person counting
- FPS monitoring
- Laptop camera support
- RTSP camera support
- Mobile camera support can be added through an IP video stream
- Web-based camera connection and disconnection
- Live detection results from the inference workflow
person-detection-web/
│
├── app.py
├── roboflow_stream.py
├── config.py
├── requirements.txt
│
├── templates/
│ └── index.html
│
├── static/
│ ├── style.css
│ └── script.js
│
└── README.md