An integrated attention monitoring system is presented. This system, including a novel head- and gaze-tracker, a lanetracker, and CAN bus integration, provides a unique opportunity to study how inattention affects driving behavior. Preliminary road tests show that the head- and gaze-tracker is robust to lighting variations, vibrations, and is reliable. Real-time data on natural driver visual behavior in real vehicles, together with real-time context recognition and prediction, provides exciting opportunities for attention assistance. Changes in visual scanning patterns and gaze fixations during driving situations that are distracting should be measurable by comparing visual behavior with vehicle performance and task actions. We also hope to guide development of human-system interfaces by learning about the visual demands of interiors and about visual behavior in general. This research should have a very positive impact on risk prevention and road safety. (Author/publisher).
Samenvatting