P‐34: Compare and Model Multi‐level Stereoscopic 3D Visual Fatigue Based on EEG

Abstract

In this paper, we designed an experiment to observe changes for different fatigue level in human brain when watching random dot stereogram (RDS) using the method of EEG. The subjects were asked to watch 3D content for 30 minutes and the disparity of RDS changed every 10 seconds. During this watching task, subjects should rate their extent of fatigue during the experiment on the scale of one to five only when they felt worse or better. 30 channels electroencephalogram (EEG) signals were recorded through the experiment. Finally, the relationship between various fatigue levels and disparities which revealed by EEG signals were investigated. Our results show that the power of waveband, namely alpha increased with the fatigue level increasing from one to four during watching RDS, but decreased in level five. We have employed logistic regression to model fatigue level and achieved classification accuracy of 94.18%. Our research has important implications for the continuous monitoring of user's fatigue in the future.

Effects of disparity on visual discomfort caused by short-term stereoscopic viewing based on electroencephalograph analysis

“…Malik et al [ 9 ] compared the EEG absolute power differences, coherence and complexity, then concluded that 3D viewing is more attractive than 2D and may cause high attention and involvement of working memory manipulations. Other researches also demonstrated the relevance between visual fatigue and disparity in stereo viewing based on EEG analysis [ 10 – 13 ]. Kang et al [ 14 ] developed a platform to facilitate comfortable stereo video viewing using EEG-based visual discomfort evaluation technology.…”