Fatigue Detection Based On Eye Behavior Analysis From Forehead Electrooculography

This study gives an approach of a new method of collecting elec-trooculography (EOG) signals for fatigue detection and using some exper-iments to prove the accuracy of this theory. Compared with traditionalEOG signals that are collected by the device placed around the eyes, thismethod only place the device on forehead to achieve excellent comfort.We also introduce the best placement of a set of electrodes, and extrac-t the same features as the traditional EOG signals for fatigue detection.Compared with previous study that makes use of four kinds of features,eye blinks, slow eye movements (SEM), rapid eye movements (REM) andenergy, this method extracts eye behavior as new features, which can char-acterize the state and behavior of eyes. This method with high precisionis fast to implement and especially convenient to collect signals, and canbe produced as an online fatigue detection algorithm. With usage of dryelectrodeamplifers, thismethodhasapromisingprospectofpracticaluse.