Projection function for driver fatigue monitoring with monocular camera

The image projection has been proven to be an effective method for extracting image features. This paper utilizes multiple facial cues extracted by the projection function for driver vigilance monitoring. To measure vigilance, we calculate two metrics: PERCLOS and PERROTAT from facial state estimation. The main advantage of the presented system in contrast to existing ones is the usage of cheap equipment (only one monocular camera) to achieve a good cost-performance ratio and fast computation time.