Automated Detection of Eye Blink and Conversion to voice using ML Algorithms

Abstract

The motive of the paper is to develop a real time method for the paralysed patients, which is the conversion of eye blinks to voice. The motive behind this research paper is to provide assistance for the disabled who always requires assistance for the interaction with hands. Hence we provide them help by fulfilling their requirement. We use certain algorithms for this need. For the facial and eye detection we use the Haar Cascade Classifier. It uses Haar like features for finding the axial positioning of the eyes. An efficient method for tracking the eye was developed which is used to detect the eye position. We also use the Eye Aspect Ratio (EAR) to find the state of the eye and estimate the opening and closing of the eyes. Basically, we use a live stream video in this project as the input and capture the Eye Blinks. This is further processed by the OpenCV Library and henceforth provides the required assistance.

 Keywords: paralysed patients, Haar cascade classifier, Eye aspect ratio (EAR), Eye blinking, OpenCV.