Abstract
In this paper, we propose an eye detection method that is adaptive to facial poses, using a particle filter and gradient directional features. To estimate the boundary between the iris and sclera or eyelid, the gradient intensities are calculated by four directional Prewitt filters in four regions. The likelihood used in the particle filter is obtained by averaging the gradient intensities for the specific direction in the four regions and the upper eyelid area. Moreover, incorrect detection is avoided by using the roll angle of the face and the eye distance, derived from the positional information of both eyes. From experimental results, the average detection rates of both eyes for roll, yaw, and pitch angles of the face is more than 90 % by using rejection function for incorrect eye detection. In addition, the rejection function provides the 2.7 %, 4.4 %, and 5.0 %increment in average detection rates of both eyes for roll, yaw and pitch facial angles, respectively. The experimental results show that the proposed eye detection method is robust to facial pose changes.