To monitor driver vigilance is a complex problem that requires an user accepted system able to measure and interpret the symptoms independently of driver's characteristics, way of driving environment and then to warn the driver about the current situation. Since more then ten years, all over the world several teams have tried to propose solutions. Unfortunately, up to now, no efficient and reliable system could reach to the market. Following the previous studies, a multi-sensorial approach is presented. Then specific sensors have been developed and improved with respect to specific needs. Nevertheless, some work remains to be done specifically on eyelid sensor. Previous projects as for example SAVE were assuming that it was possible to identify in addition to the normal behavior several classes of impairment: drunk, sleepy. Although this approach looks interesting in a research context but this is not realistic in an applicative one (it is not possible to ask a driver to drive in impaired conditions in order to learn its behavior). Four information were selected: (1) the vehicle speed (provided by the engine control unit) and (2) the steering wheel movement (resistive or electromagnetic sensor). The two last information are based on video processing and are concerned with specific developments: (1) eyelid sensor and (2) vehicle lateral position. Thus a one-class pattern recognition (normal behavior) approach was promoted based on statistical classification that looks much more adapted to the actual problem. At last specific developments have been achieved in order to identify discriminatory parameters to provide the diagnosis unit. Current results and orientation look promising. Nevertheless they must be confirmed in a near future by intensive tests.
Driver vigilance monitoring - new developments
Neue Entwicklungen auf dem Gebiet der Fahrerwachsamkeitsüberwachung
2002
6 Seiten, 10 Bilder, 15 Quellen
Aufsatz (Konferenz)
Datenträger
Englisch
Driver Vigilance Monitoring, A New Appoach
British Library Conference Proceedings | 2003
|Driver vigilance monitoring, a new approach
IEEE | 2002
|Smart video sensor for driver vigilance monitoring
Tema Archiv | 1998
|Real-time system for monitoring driver vigilance
IEEE | 2004
|Real-time system for monitoring driver vigilance
IEEE | 2006
|