A Text Input System Based on Information of Voluntary Blink and Eye-Gaze Using an Image Analysis

Abstract

Eye-gaze input interfaces have been proposed. A main purpose of these systems is communication aid for the severely disabled. We have previously developed a text input system using eye-gaze based on an image analysis under natural light. This input system is employed dwell time based method to select indicators. By employing voluntary (conscious) blink based method rather than dwell time as indicator selection, it is expected to increase the efficiency of text input. We propose a new text input system based on eye-gaze and voluntary blink method. The newly added eye blink measurement function is necessary to realize high time resolution. Furthermore, each of the eye-gaze and eye blink measurements is very time consuming. If these measurements are performed sequentially, it is difficult to finish calculation within required time for real-time processing. To cope with this problem, we realize the concurrent processing of these measurements by multi-thread structure. We conducted an examination of nine subjects to input Japanese text using our new input system. All of nine subjects could complete the task of Japanese text input. Based on a result of this experiment, we evaluated our new system.