2A2-G06 瞳位置検出素子の最適配置および瞳位置検出アルゴリズム(感覚・運動・計測(2))

Abstract

A wide variety of applications featuring eye-gaze tracking systems are expected, such as in communication technology and welfare field. We set dye-sensitized photovoltaic cells on the glasses and determine the pupil position by detecting potential difference from them. In this paper, we make a program that can judge whether the cell positions we set are optimized and make an algorithm to detect horizontal and vertical pupil position by using the least squared method and Newton-Raphson method. With the use of this system, we achieved the line of sight (LOS) detection accuracy of ±2.9° under constant illumination.