While the dangers of driving while distracted by cell phones has been grabbing , the U.S. National Highway Traffic Safety Administration says at least 100,000 road crashes a year are caused by driver fatigue.

A paper (A non-rigid motion estimation algorithm for yawn detection in human drivers) describing the new yawn detector is included in the International Journal of Computational Vision and Robotics. The authors hail from Vanderbilt University in Nashville as well as ITER in Orissa, India and UUT if West Bengal, India.

They describe an in-vehicle camera that uses image-processing software and an algorithm to grab images of a driver’s face and take notice of yawning vs. other facial movements such as smiling or singing along to that . A warning system can be triggered when yawning frequency indicates fatigue.

Such a system, the researchers say, could be far less expensive and intrusive than a system that involved hooking a driver up to detect such other fatigue indicators as brain waves and heart rate.