Development of an integrated wearable EOG-based eye-tracker

Type: Master thesis
Students: Roberto Celli, Christian Meyer
Advisor: Andreas Bulling, Jamie Ward
Project: Eye Movements and Visual Perception

In this thesis the design, implementation and test of a EOG recording device is described.
The device had to meet several criteria such as being wearable, integrated, long running and capable of processing signals online. Dry Ag/AgCl electrodes were to be used with the system and state of the art biopotential measuring was investigated for the acquisition of the EOG signals. An accelerometer as well as a light sensor were incorporated into the device, in order to strengthen the EOG signal integrity. A Digital Signal Processor (DSP) was programmed to process the sampled signals and to either send them directly to a computer via Bluetooth or to store them in a Multi Media Card (MMC). Real time EOG signal processing was implemented on a computer. For this purpose, existing algorithms were rewritten to meet online processing demands. An EOG Viewer application was created which allows to perform real time signal analysis on a stream of data. As a proof that the complete system works as intended, an experiment was conducted with several subjects. The experiment consisted of a game which the subject had to play by moving his/her eyes.