Abstract
In the field of active safety, driver monitoring is one of the most important challenges because most of the traffic accidents have been caused by drivers mental state. Some studies for evaluating mental state during driving have been reported in literature; however drivers mental state should be evaluated quantitatively in order to develop better and safer systemsfor the future. We propose an estimation method of drivers mental workload quantitatively using physiological information. This paper describes good estimation results of several mental states using drivers visual information.