Eye movement analysis is useful in many applications such as security systems, HCI(human computer interface), visual telephone communication, and systems for manmachine interaction. Also application areas include human factors research and human performance evaluation.
Over the past two decades many computer-based systems have been developed for the study of eye movements. It basically consists of two stages, namely, tracking the eye features and analyzing the pattern of eye dynamics. Eye feature tracking usually refers to the estimation of the eye pointing direction, either within the eye socket, or with respect to some scene being viewed.
The main objective of this paper is to propose an efficient approach for real-time eye gaze tracking from image sequence and magnetic sensory information. In the proposed system, three basis images that are extracted by preprocessing an eye image are chosen as tracking parameters. To this end, we formulate an efficient model of the eye, using the gray level information of the eye image sequence. This image-based method is different from traditionally eye movement measuring systems, which are used by model-based algorithms, and is implemented using an inexpensive PC, fitted with a commercial-grade video camera and image frame grabber. And we use magnetic sensory information to track a head motion. The inputs to the system are images taken by a camera and datum of a magnetic sensor. The measuring data are sufficient to describe eye and head movement, because the camera and the receiver of a magnetic sensor are stationary with respect to the head. Experimental result shows the validity of the proposed system, especially in real time application aspect, and also shows the feasibility of using new computer interface instead of the mouse.
