Applications based on eye-blink detection have increased, which requires eye-blink detection to be robust and non-
intrusive irrespective of the changes in the user’s facial pose. However, most previous studies of camera based blink
detection have the disadvantage that the performances were affected by the facial pose. They also focused on blink
detection using only frontal facial images. To overcome them, we propose a new method of blink detection robust to
the facial pose.
This research is novel in the following four ways compared to previous studies. First, the face and eye regions are
detected by using both AdaBoost face detector and Lucas-Kanade-Tomasi (LKT) based method in order to be robust
to facial pose. Second, the state of the eye being open or closed needed for blink detection is determined based on
two features, the ratio of the height to the width of the eye region in a still image, and the cumulative difference of the
number of black pixels of the eye region using an adaptive threshold in successive images. These two features are
robustly extracted irrespective of the illumination variations by using illumination normalization. Third, the accuracy of
determining eye state, open or close is increased by combining the above two features based on the support vector
machine (SVM). Finally, the SVM classifier for determining the eye state is adaptively selected according to the facial
rotation.
Experimental results with three databases showed that the blink detection by the proposed method was robust to
various facial poses.