Distracted driving contributes to a significant portion of vehicle accidents and deaths. To mitigate distracted driving, the University of Virginia Center to Promote Effective Youth Development sponsored the development of a noninvasive system to detect and warn drivers of their distraction. This distracted driving warning system is installed in the medium-fidelity driving simulator of the Virginia Driving Safety Laboratory (VDSL). The system is composed of (i) the Microsoft Kinect motion sensing hardware for tracking head and skeletal movements and (ii) a custom software application for identifying four distractions and outputting audio alerts. The system is able to identify (a) reaching for a moving object, (b) talking on a cell phone, (c) personal hygiene, and (d) looking at an external object. To recognize these distractions, the algorithms for (a), (b), and (c) use the relative distances between spatial locations of various skeletal joints, and the algorithm for (d) use the yaw, pitch and roll of the head. The system deems the driver distracted if any of these gestures are sustained for more than two seconds. When the driver is deemed distracted, the system produces audio signals that increase in frequency as the time of the distraction increases, alerting the drivers of their distraction. The distracted driving warning system is tested with three participants performing distracted behaviors while driving the VDSL simulator. The system correctly identifies (a) at 100%, (b) at 33%, (c) at 50%, and (d) at 66%. These success rates show the feasibility of employing the Kinect to identify driver distraction. However, the system can improve with refining motion capture and eye tracking technology for some complex distraction behaviors. The application of commercially available motion capture technology appears promising for studying and monitoring driver behavior related to distraction.
[1]
R. Barkley,et al.
Motor vehicle driving competencies and risks in teens and young adults with attention deficit hyperactivity disorder.
,
1996,
Pediatrics.
[2]
Thomas A. Dingus,et al.
An overview of the 100-car naturalistic study and findings
,
2005
.
[4]
Pujitha Gunaratne,et al.
A closed-loop crash warning system based on heterogeneous cues
,
2011,
The 5th International Conference on Automation, Robotics and Applications.
[5]
Thomas A. Dingus,et al.
The Impact of Driver Inattention on Near-Crash/Crash Risk: An Analysis Using the 100-Car Naturalistic Driving Study Data
,
2006
.
[6]
John D. Lee,et al.
How Dangerous Is Looking Away From the Road? Algorithms Predict Crash Risk From Glance Patterns in Naturalistic Driving
,
2012,
Hum. Factors.
[7]
M. Trivedi,et al.
Head and eye gaze dynamics during visual attention shifts in complex environments.
,
2012,
Journal of vision.
[8]
D. Cox,et al.
Effect of stimulant medication on driving performance of young adults with attention-deficit hyperactivity disorder: a preliminary double-blind placebo controlled trial.
,
2000,
The Journal of nervous and mental disease.