Detection and Analysis of Drowsiness in Human Beings Using Multimodal Signals

The state of drowsiness can be characterised as an intermediate state of mind which occurs between the alert state and the sleep state. The alertness of the mind is reflected immediately through the sense organs and other body parts. Automatic detection and analysis of drowsy state of mind is essential in applications where the human's mental status is important. One such scenario is monitoring driver's alertness while he is driving. A multi-modal approach is analysed for detecting the drowsy state in humans. Two modalities are considered here, video information and bio signals, for analysis. Visual information conveys a lot about the human alertness. The precise indicators from the video information need to be identified and captured for analysis and detection. The bio signal that indicates human brain alertness is EEG signal. The physical and mental alertness are analysed for detecting drowsiness state of a human being. A framework is proposed for drowsiness detection of humans in real-time.