Annu Int Conf IEEE Eng Med Biol Soc. 2017 Jul;2017:1509-1512. doi: 10.1109/EMBC.2017.8037122.
The purpose of this study is to analyze eye blinking patterns using high-speed camera without additional instructions and illuminations. The experimental video clips are taken at 240 frames per second, which is twice the flickering rate of room light. The eye blinking sequences are segmented from the whole video clip after reducing noise and baseline wander. The shapes and positions of the upper eyelid in blinking sequences are evaluated by a polynomial curve fitting algorithm and visualized to the eyelid position graph. The graph made it easy to recognize the eye blinking patterns. In order to analyze patterns, eye blinking parameters are calculated, and an eye blinking cycle is divided into three phases as Closing phase', Closed phase', and `Opening phase'. In the experiments with forty volunteers, proposed method can analyze eye blinking patterns qualitatively.
