Home > Published Issues > 2015 > Volume 3, No. 2, December 2015 >

Driver Attention and Behavior Detection with Kinect

Cleshain Solomon and Zenghui Wang
Department of Electrical and Mining Engineering, University of South Africa, Florida 1710, South Africa

Abstract—Modern vehicles are designed to protect occupants in the event of a crash. However, passenger protection can be combined with collision avoidance. Statistics have shown that human error is the number one contributor to road accidents. Advanced driver behavior and monitoring systems have been developed by manufacturers in recent years and many have been proven to be effective systems in the prevention of accidents. However, these systems do not provide the complete solution and the systems only detect driver fatigue but do not offer gesture detection. In this paper, a system that detects driver fatigue and distraction has been developed using non-invasive machine vision concepts to monitor observable driver behavior. Moreover, cellular telephone detection was also considered. This paper also explores how driver error can be reduced through inexpensive technology (a Windows developer Toolkit Kinect) which observes driver behavior and reacts when certain unwanted patterns in behavior have been recognized. The experiments have been done to validate the efficiency of this new system.

Index Terms—Kinect, driver fatigue, driver behavior, behavior detection

Cite: Cleshain Solomon and Zenghui Wang, "Driver Attention and Behavior Detection with Kinect," Journal of Image and Graphics, Vol. 3, No. 2, pp. 84-89, December 2015. doi: 10.18178/joig.3.2.84-89