Preliminary Study On EEG-based Analysis Of Discomfort Caused By Watching3D Images

With the development of computer graphics technology, the stereoscopic display technology creates a virtual real-world, which could have an immersed sense on viewers in the virtual environment. This technology becomes a new hot spot in the technical fields and will be widely applied in other fields. A large number of3D-films projections also attract more and more attention on stereoscopic technology. Stereoscopic technology represents a technical innovation and provides with strong sense of stereo and immersion, however, if having seen a3D film for a long time (more than30minutes), viewers are likely to feel headache, visual fatigue, nausea and other uncomfortable symptoms, which, to some extent, affects the application and development of stereoscopic technology. At present, the analysis of the neurophysiological changes in the3D viewing conditions and discomfort caused by3D images has been lacking. Therefore, this topic proposes to analyze the subjective relations between the discomfort caused by3D images and EEG characteristics through the study of the changes of the EEG characteristics in3D viewing conditions. This study will contribute to the in-depth research on the discomfort caused by3D images and lay foundation for the improvement of stereoscopic technology.EEG is the electrical phenomena demonstrated by the overall activity of brain cells in the cerebral cortex or the surface of the scalp. It is a comprehensive reflection of the physiological activity of the brain. EEG could be adopted to detect the brain’s thinking, pathology, cognition and other central neurons’activity.In order to explore the EEG information reflected by the discomfort caused by3D images, EEG experiments in2D and3D viewing mode were implemented in this study. In the experiment, the viewer’s EEG was recorded. Also, we made the scale survey before the viewing, in the viewing process and after the short break at the end of the viewing.Firstly, the EEG signals acquired in this study were preprocessed. Subsequently, the preprocessed signals were studied using the spectral feature analysis and nonlinear analysis. The spectral characteristics were comprised of the average power spectrum and the ratio of the relative power spectrum. In the nonlinear analysis, we chose the parameters such as the entropy, KC complexity and approximate entropy. Also, Statistical tests were used to explore the differences between the2D and3D experimental conditions. The analysis on the viewers’scale survey showed that there were a significant increase of discomfort in the3D viewing condition, and the discomfort is mainly concentrated in the visual sense and limb tension. The spectral analysis showed that in the3D viewing condition, the alpha rhythm in EEG gradually reduced while beta rhythm increased, therefore, the R ratio (beta/alpha) increased during the viewing. The analysis of nonlinear characteristics showed that in3D viewing condition, the entropy, KC complexity and approximate entropy in frontal lobe represented significant differences. This indicated that in3D condition, the viewers will show a stronger neuronal excitability, and thus more likely to induce visual fatigue. Besides, KC complexity and approximate entropy in parietal and temporal lobes also showed significant changes, which indicated the transfer rate and the physiological state changes.The results above showed that there were significant differences between2D and3D viewing conditions in the subjective scales, EEG spectral and nonlinear characteristic analysis. Therefore, the study on discomfort by3D images with electrophysiological signals is worth further study.