This paper improves saccadic movement detection by quantifying
these movements. This is achieved by taking the findings of
Barea et al. (2002), in which an electrooculographic eye model with
an eye movements detector based on a linear saccadic eye model
was used, and improving the results by increasing detection accuracy
by employing a wavelet and avoiding EOG signal degradation
over time by using a neural network to identify valid saccadic
movements.