An Algorithm for Detecting the Center of the Iris Using a Color Image and Improvements of Its Processing Speed

Abstract

We have studied a man-machine interface system by using an eye gaze detection technique. Target users for this system are ordinary people and orthopedically impaired people who have no movement in their hands and heads. Ordinary people can use this system under the condition that they do not move their heads. An algorithm for detecting the center of the iris is proposed in this paper. An image of the eye is captured with a small video camera and the center of the iris is obtained by filtering in color space and pattern matching with a ringshaped template. Improvement of the pattern matching speed is also described and the effectiveness of the reduction of calculation costs is shown in the paper. Furthermore, the accuracy of the eye gaze detection was evaluated.