Facial recognition system for driver vigilance monitoring

Many automobile accidents are related to drivers lacking required levels of vigilance to properly control their vehicles. In this paper we present a system that monitors the activity of parts of the face, in particular the eyes, in order to predict expressions of somnolence. The input to the system is a sequence of images of the face of a car driver, captured by a video camera. The system makes an assessment based on the movement and position of the eyes and eyelids. The system is tested in a car simulation environment. The results are presented.