Abstract
This paper describes the design of a low-cost assistive communication prototype in which voluntary eye blinks are sensed by a reflective infrared (IR) sensor on an eyeglass-type frame and converted into Morse code. A TCRT5000 sensor held about 10–15 mm from the dominant eye feeds an Arduino Uno, which classifies each blink as a dot or a dash, decodes the characters, and shows them on an LCD, with serial output for text-to-speech. Unlike EOG systems and camera-based trackers, which need skin-contact electrodes and cameras facilitating image processing, the current design uses a contactless IR sensor and a microcontroller interfaced to a smartphone. Ten healthy volunteers (aged 22–58 years, no prior Morse experience) took part in one training session and three test sessions. Mean accuracy increased from 87.3% to 95.6% (overall 94.7 ± 3.2%), and throughput increased from 5.2 to 9.8 characters per minute. Fatigue-curve studies showed that prolonged blinking over a 20 min session decreased the accuracy from 95.6% to 86.8%; speed decreased from 12.0 to 9.0 characters per minute, and self-rated eye tiredness (0–10 scale) rose from 1.0 to 5.2. The current prototype can be applicable as an alternative communication to speech.