Real-Time Eye Detection using Face Circle Fitting and

Dark-Pixel Filtering*


Daw-Tung Lin and Chian-Ming Yang

Department of Computer Science and Information Engineering

Chung-Hua University

No. 707, Sec. 2, Wu-Fu Rd., Hsinchu, Taiwan, 30012



In this paper, we develop a real time system for eye detection and tracking in video sequences. The HSI color model is used to extract skin-color pixels. The same type of pixels are gathered by a region-growing algorithm and the maximum area of this region is selected as the face candidate. Furthermore, a Face-Circle Fitting (FCF) method is proposed to confirm the exacted face region. Once the face is detected, a Dark-Pixel Filter (DPF) is used to perform the eye detection and tracking. The proposed system has been implemented on personal computer with a PC-camera and can perform eye detection and tracking in real-time. The correct eye identification rate is as high as 92% during the actual operation.