부산대학교 도서관

The ability to identify the location of human pupil offers us the possibility to know the eye gaze target of a person. This infers the possibility to create new methods of human-machine interaction. Nevertheless there are some barriers and restrictions that are shown in the integrations, as intrusion, robustness, availability and price of the eye track systems. This paper describes the develop and usability of an eye gaze input system using a web cam with infrared LEDs over a head mounting system. Most available eye gazing with head mounting have two characteristic that hinder being widely used as human computer interfaces. The first is represented by the need of a calibration stage before the use; the second is about the low tolerance for head movements during all the using time. This work solved that problem using a 3 axis accelerometer as head tilt sensor to make a compensation. Our main objective is to create a communication system for people with motor disabilities based exclusively in self-inflicted ocular movement.