Technical Report 2007-12-01
Internetworking and Media Communications Research Laboratories
Department of Computer Science, Kent State University


Dissertation Title: 

Eye Movement Prediction by Oculomotor Plant Modeling with Kalman Filter

Oleg V. Komogortsev

Advisor: Prof. Javed I. Khan
Department of Computer Science
Kent State University

December 2007


In this Dissertation, the Oculomotor Plant Kalman Filter (OPKF) framework is designed. The main goal of the OPKF is to predict future eye movement trajectories. Within the OPKF a Kalman Filter is used as the base eye movement prediction mechanism. Additionally, an Oculomotor Plant Mechanical Model (OPMM) is integrated into the Kalman Filter to improve the accuracy of prediction. The OPMM takes the following into account anatomical properties of the eye: muscle location, elasticity, viscosity, eye-globe inertia, muscle active state tension, length tension and force velocity relationship. The OPKF framework is capable of detecting eye movement types and of maintaining an eye movement prediction signal during eye tracking failures. The accuracy of the eye movement prediction achieved by the OPKF is compared to several other prediction models with simulation results indicating the superiority of the OPKF framework. The practical use of the OPKF lies in the Human Computer Interaction domain, specifically in the areas of direct eye gaze input and interactive displays.


Last Modified: Sep 2007.