A calibrated combined head-eye tracking system

Real-time gaze tracking is a promising interaction technique for virtual environments (VE). To determine the direction of gaze in space both the head pose and the orientation of the eyes in the head need to be estimated. Most head-eye movement measurement techniques are of limited use during free body and head motion. Although some methods have been proposed to overcome this problem, limitations still exist. An improved head-eye tracking system is proposed and developed for both desktop and immersive HCI (human-computer interaction) applications with free body and head motion. This system is based upon a video-based head-mounted eye tracking system and a hybrid acoustic-inertial head tracking system. Because the system is able to measure the point of regard in a scene during relatively large head movement, it can be used in immersive projection based VE systems like the CAVE™. The calibration of the head-eye tracking system is one of the most important issues that need to be addressed. In this thesis, a simple boresight-like calibration method is proposed and evaluated. The performance of each subsystem as well as the whole system are also evaluated and discussed in this thesis.