- 期刊
An Eye-Tracking Human-Machine Interface Using an Auto Correction Method
並列摘要
In this paper, a novel design of an eye-tracking device with an auto correction method is discussed. In this eye-tracking system, an infrared CCD camera captures the eye image from the screen with the help of a USB interface. The locations of the pupil of eye image are calculated, compared with the previous point and mapped to the point of the screen. When the cursor is out of control of the eye, a correction theory is used to account for the reflexive action of psychological natural reaction to correct the system automatically. The theoretical design of automatic correction can continue until the unreasonable action of the eye is judged to be ignored. Experimental results verified that this image system provides an economical and effective method for eye-tracking. With the auto correction method, the gaze point can be extracted successfully by means of the correction for incorrect mapping ratio with 97% accuracy. We hope this auto correction method for the eye-control functions can help some people to use computers more conveniently.