Iris communicator

Abstract

This research purposes an eye pupil tracking experimental software to enable communication abilities for individuals with severe motor impairment. Since the ability of moving eye pupil is often preserved, even in severe conditions such as locked-in syndrome, Iris Communicator allows text entry with eye movement as the only input modality. To begin with, a camera was used to get video input from where we got BGR (Blue-Green-Red) images. Afterwards, we converted those BGR images into grayscale images. Then, we used Adaptive Threshold method for smoothing the images. The images were finally ready for eye pupil detection. Here, we used Hough Circle Transform method for detecting eye pupil as a circular object. Then, we programmed an algorithm to plot and keep track of the eye pupil movement and we designed a new keyboard layout to synchronize. In addition, we programmed a calibration method to extend the gaze area of eye pupil. Finally, we included text to speech feature by using pyTTSx3 and created a combination chart for each English letter. In future, we will be working on creating a user interface using Tkinter. Also, we will add more features like taking blink input and adding a cursor mode.