Complexity Analysis of Ocular Signal for Detection of Human Fatigue Using Small Datasets

Abstract

The electrooculogram, often known as an EOG signal, can frequently be used to quantify the amount of drowsiness and exhaustion that a person is experiencing in their body. Electrooculogram signal analysis is primarily a process that does not include any invasive procedures and is used to evaluate the movement of the eye in both horizontal as well as in the vertical direction. The primary purpose of this research is to identify signs of exhaustion using the process of complexity analysis using non-invasive data collected from twenty people who are deemed to be in good physical and mental health. This investigation makes use of three distinct visual signals at three distinct frequencies, and it also considers three distinct time slots during the day. Based on real-time EOG signals that have been gathered in a non-invasive manner, oculo-parameters such as moving window fuzzy entropy and moving window dispersion entropy can be computed. According to the findings of the experiment, the level of complexity of both horizontal and vertical data steadily grows throughout the afternoon and evening as cue frequency steadily rises and we can ascertain the fatigue of the body. Drowsiness and exhaustion in humans may be caused by factors such as muscle stress, insufficient sleep, and a significant time gap between two different sleep cycles.