Face detection and tracking

Real-time face tracking with Viola-Jones detection and Arduino-controlled pan and tilt servos.

NIT Rourkela, Aug 2018 to Dec 2018 (Product Development Lab)

A real-time system that follows a person’s face with a regular webcam and two servo motors.

  • Detection. MATLAB grabs 320×240 frames from the webcam and finds the face with a Viola-Jones cascade detector.
  • Control. The center of the detected face is sent over a serial link to an Arduino, which compares it with the middle of the frame and steps a pan servo and a tilt servo to follow the face.
  • Tracking. Once a face is found, the system can switch to histogram-based CAMShift tracking on the hue channel, which follows the face without running the detector on every frame.

Code on GitHub