EOG based vigilance monitoring system

EOG signal is being used as the diagnostic tool forvarious neurological disorders. While acquiring the EOG signal, blink artifacts and ocular artifacts are unavoidable. Though these artifacts corrupt the EOG signal, blinks provide useful information about the vigilance status of the subjects who are engaged in activities such as driving, reading etc. Blink detection therefore plays an important role in assessing the efficiency of the people at workplace. Most significantly it saves many lives by preventing the accidents caused by fatigued drivers driving for long durations. To address this issue, a lab view based EOG recorder has been developed to acquire the EOG data and subsequently analyze it in MATLAB. EOG data is acquired from 20 subjects under active and drowsy conditions. The subjects were chosen in such a way that they are deprived of sleep for at least 18 hours for the drowsiness condition. The Data acquisition for the active condition is performed early in the morning with consent from the subjects. The subjects were asked to perform an activity while acquiring the data. Video recording is also performed simultaneously to verify the performance of the EOG system. An algorithm is developed to analyze the raw EOG data so as to detect the blink and determine its parameters. The parameters considered in the present work are blink duration, blink rate, eye opening time, eye closing time, and the peak velocity. For most of the subjects, these blink parameters have shown significant difference in active and drowsy conditions. Therefore the system developed during this work would be useful to detect the drowsiness of the driver and prevent many accidents thereby saving many precious lives.