Monitoring drivers drowsiness using a wide angle lens

This paper proposes a robust and nonintrusive system for monitoring drivers drowsiness using a wide-angle and FishEye lenses. The proposed scheme begins by extracting the face from the video frame using face detector based on skin color. Then the proposed system using wide-angle lens allows detecting drowsiness by extracting micro-sleeps using the presence of the iris in the eyes from frontal and profile faces, the eye state analysis is based on Circular Hough Transform (CHT) and is applied on eyes extracted regions. In this work, we present the effect of wide-angle and FishEye lenses on driver imaging, nearby driver appears very large and objects at a moderate distance appear small and far away. In one side, this exaggeration of relative size can be used to make foreground objects more prominent and striking in quality and intensity informations, while capturing expansive backgrounds. In the other side, the wide-angle lens allows us to detect the state of the eye even if the driver is moderate in profile, which is not the case for conventional lenses. The system was tested with different sequences recorded in various conditions, with different subjects. These sequences were captured in both normal car and intelligent vehicle named `SeTcar'. Some experimental results about the performance of the system are presented.