The system proposed in this paper is able to identify movements
of ±10 and greater with significant accuracy (2 degree error).
In other words, with this system it is possible to detect
complex sequences of eye movements (in accordance with users’
capabilities). Superimposing a grid on the graphical interface reveals
the system is able to identify with significant accuracy 9
movements on the horizontal axis and 7 on the vertical axis, as
well as their corresponding diagonals (taking into account that
users may make movements of ±40 on the horizontal axis and
±30 on the vertical axis).
T