Develop and test a wearable eye tracker

Ong, Chia Koon (2020) Develop and test a wearable eye tracker. Final Year Project, UTAR.

Preview

PDF Download (3786Kb) | Preview

Abstract

Eye tracking is a technology that tracks the pupil of the user’s eyes and estimates her/his gaze direction to realize gaze control computer and machine interface. This can be useful for patients with severe motor disabilities to control a motorized wheelchair. However, most of the available wearable eye trackers currently suffer accuracy issue over time of usage due to slippage and drifting problems. These problems are especially significant and critical when it is used with on automated wheelchair moving on uneven surfaces and setting up vibrations. Therefore, a wearable eye tracker with robust frame design is necessary to overcome the challenges. The accuracy of the eye tracker is evaluated and compared before and after a series of designed conditioned that are similar to habitual behaviours of user in real life application which contributes to the slippage of an eye frame. The results of accuracy and precision were obtained in the factor of deviation and RMS (root-mean-square) respectively in regard to the calibrated result. Although it was still suffering from the effect of frame movement especially in the z-direction, the overall improvement on the frame was significant compared to previous research performed with commercial wearable eye tracker.

Actions (login required)