3D auto-calibration method for head-mounted binocular gaze tracker as human-robot interface

Authors:
Su Hyun Kwon;Min Young Kim
Affiliations:
Kyungpook National University, Daegu, South Korea;Kyungpook National University, Daegu, South Korea
Venue:
Proceedings of the 8th ACM/IEEE international conference on Human-robot interaction
Year:
2013

Citing 1
Cited 0

Starburst: A hybrid algorithm for video-based eye tracking combining feature-based and model-based approaches

CVPR '05 Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) - Workshops - Volume 03

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper presents a novel calibration method for a head-mounted binocular gaze tracker that enables the human gaze point, representing the selective visual attention of the user, to be tracked in 3D space. The proposed method utilizes two calibration planes with visual marks to calculate the mapping points between a forward-looking camera and two eye-monitoring cameras in an expanded 3D spatial domain. As a result, the visually attentive point of the user can be tracked, regardless of variations in the distance from the user to the target object. The proposed method also provides a more convenient calibration procedure and more accurate results in tests than the previous method suggested by the authors. The performance is tested when varying the 3D position of an attentive object, and the experimental results are discussed.