Development of a head-mounted eye tracker for human-computer interaction

Ong, Li Kang (2024) Development of a head-mounted eye tracker for human-computer interaction. Final Year Project, UTAR.

Preview

PDF Download (2923Kb) | Preview

Abstract

This paper outlines the design, creation, and integration of an affordable head�mounted eye tracker system specifically designed for applications involving human-computer interaction (HCI). The technology employs easily accessible webcams and 3D-printed parts to create a lightweight and cost-effective eye�tracking gadget. The gaze detection software framework utilizes the Pupil Capture program, which has been developed by Pupil Labs. This ensures precise and dependable eye tracking. An important contribution of this research is the adaptation and integration of a cursor control plugin, which improves the usability of the eye tracker for human-computer interaction tasks. By including the cursor control plugin, users are able to engage with graphical user interfaces and manipulate the mouse cursor exclusively by using eye movements. This enables a seamless and instinctive engagement with computers. In addition, a series of comprehensive data-gathering experiments were carried out to assess the performance and efficacy of the eye tracker system that was built. The experiments included a range of scenarios and tasks that are commonly found in HCI applications. Eye-tracking capabilities in terms of its accuracy, precision, and robustness were evaluated by the application dart board graphical user interface. Subsequently, the gazed-controlled mouse cursor could be used for typing in a virtual keyboard. Keywords: Eye-tracking, Human-computer interaction, Head-mounted, Gaze detection, Cursor control, Pupil Labs, Virtual Keyboard.

Actions (login required)