Blink-to-Speech Communication System (Graduation Project)
A portable Raspberry Pi and OpenCV based image-processing device that turns the blinks of ALS patients into meaningful messages.
For my graduation project I built a portable device that lets ALS patients — who have lost speech and motor abilities — communicate with the people around them. The system tracks the eye region in the camera feed and maps blink sequences to predefined messages.
Technical approach
- Face and eye region detection with OpenCV on a Raspberry Pi
- Command parsing based on blink duration and frequency
- Time-window filtering to reduce false positives
- Displaying the selected message on screen
The aim was a practical, low-cost hardware answer to an accessibility problem.