A novel feature based rapid eye state detection method
Feng Wang, Mi Zhou, Bingchu Zhu · 2009
A rapid algorithm for eye state detection is proposed in this paper. It takes use of gray scale characteristics of the eye and upper eyelid bending direction to distinguish the open and closed eye states. Circle similarity of region is proposed as an index to characterize the eye state. And a simple method based on the relation of curve and the straight line linking the two endpoints of curve is proposed to determine the upper eyelid bending direction. The algorithm is evaluated on BioID face database, Yale face database, and part of CAS-PEAL face database. The experimental results show that it achieves more than 85% accuracy of eye state detection in less than 5 ms runtime in average.