Analysis of eye movement during generation of a trajectory using human upper limb.

N. Toyoda,R. Yamamoto,J. Huang,T. Tabuta
DOI: https://doi.org/10.1109/ROBIO.2010.5723308
2010-01-01
Abstract:This paper describes an experimental analysis of eye movement during generation of a trajectory using a human upper limb, carried out with the aim of elucidating the human mechanism for visual-information recognition. In the experiments, subjects were asked to draw trajectories of six kinds of images, a complete circle, semicircle, stippled circle, tilted square, square with some missing parts, and circle with a fake center point. Reflective markers were attached to the subjects for motion capture, and a head-mounted eye-mark recorder was used to record eye movement. The results show that subjects adopted an eye-movement pattern termed the subgoal travel method (fixating in the vicinity of the fingertip while drawing/tracing) for the complete circle. Subsequently, when the target trajectory was one of the incomplete shapes, some subjects continued following the subgoal travel method by drawing the missing part of the target trajectory, and some others followed two other eye-movement patterns. One is termed a center-point fixation method, in which subjects considered the center point as the most important point for generating the target-image trajectory, and therefore, they fixed their gaze on the center throughout the experiment. The second pattern is termed a dual-point travel method, in which the subjects' gaze shifted between the center point of a displayed image and their fingertip. These results show that the center-point fixation method provides more visual information than the subgoal travel method. Further, from the results, we predict that the eye-movement pattern, movement accuracy, and drawing speed are correlated.
What problem does this paper attempt to address?