Measurement of Viewing Distances and Angles for Eye Tracking Under Different Lighting Conditions

Eye tracking can be used for eye-gaze-based authentication in public settings, such as for registration on personal computers and automated teller machines. In this study, we conducted a series of measurements with low-cost eye-tracking devices to assess the feasibility of their use in such settings. We investigated the devices' minimum and maximum viewing distance limits as well as viewing angle range. Both these parameters were tested under three different lighting conditions. The eye-tracking devices used in the measurements were among the most cost-effective devices commercially available today (Tobii EyeX (c) and Eye Tribe (c)). The results showed that the minimum and maximum viewing limit of the Tobii EyeX (c) did not vary significantly with different lighting conditions, and neither did the viewing angle range. The maximum viewing distance of the Eye Tribe (c) was significantly different between two conditions of artificial lighting, although the average difference was just 1 cm. The lowest viewing angle that allowed registration into the Eye Tribe (c) system also differed significantly between the two artificial lighting conditions, but here too the average angle difference was small (0.63 degrees). Taken together, the results show that the performance of the two low-cost eye trackers tested here is robust enough under different conditions of illuminance and luminance for registration into the system.