Real-Time Eye Movement Detection for Vision-Based Interfaces

In today's world, computers are essential for communication, education, business, healthcare, and technical innovation, all of which improve productivity, connectedness, and general quality of life. They play a vital role in promoting and supporting human endeavors in a variety of fields. For those with impairments, using computers can be quite difficult because of a variety of physical, sensory, and cognitive limitations. For people with physical disabilities, such as those who have restricted mobility or motor impairments, vision-based interfaces offer vital solutions that let them communicate with technology in ways that are not possible with traditional input techniques. This study project fosters equality of opportunity and inclusivity by making sure that technology is available to all people, irrespective of their physical capabilities. This endeavor aims to compensate people, especially those who are physically unable to operate a mouse because of their disabilities. Without the need for further hardware, users can click and move the pointer using the suggested interface merely by looking at it. In this work, the face and eye are detected using a Haar-Cascade classifier. Twenty individuals participated in the study test. The subject's face and eyes are captured and mouse events like left and right clicks and the cursor movements are mapped to the eye movements. The experimental sections showcase the results of the experiments. Each subject's accuracy will be computed. The right-click event yields a greater accuracy of 85.7% in this work.