Face Scrambler uses facial recognition and image processing to blur faces in real-time video streams, protecting the privacy of individuals captured on camera while maintaining the integrity of the video. Built with OpenCV and the face_recognition library.
How It Works
The system captures live video from a webcam and processes each frame to detect faces. Detected faces are compared against a pre-defined dataset of known individuals.
If a recognized face matches a name in the protected list, OpenCV's blur function is applied to that region. Otherwise, a rectangle is drawn with the person's name. The system also logs attendance for recognized individuals in a CSV file.
Future Directions
- Object recognition — extend privacy protection beyond faces to objects.
- Multi-camera support — deploy across multiple camera feeds simultaneously.
- Security system integration — plug into existing surveillance infrastructure.