Real-time Estimation System of the Gaze Position Based on an Electrooculogram

Xiu Zhang,Takenao Sugi,Xingyu Wang,Masatoshi Nakamura
DOI: https://doi.org/10.1007/s10015-009-0649-2
2009-01-01
Artificial Life and Robotics
Abstract:This article describes an online eye-tracking method based on an electrooculogram (EOG) to estimate gaze position. The objective is to use a biomedical signal EOG as the input of a human-machine interface for both disabled and healthy people. In this study, features of horizontal and vertical EOG signals were extracted to estimate the gaze position. Compensation for time-shift and nonlinearity were applied to improve the performance of the estimation. The estimated locus of the moving eye was compared with the locus of the target, and the deviation was about 3 cm. The results can be applied in some raw tracking fields, such as online communication, EOG-based pointers, and so on.
What problem does this paper attempt to address?