Research Of Convergence And Accommodation Conflict Theory And Perception Based On EEG

With the continuous development and improvement of science and technology,stereoscopic display technology is more and more widely used in various fields of people's life and work.However,since the current stereoscopic display technology is mostly based on stereo disparity,the conflict of convergence adjustment caused by parallax causes a phenomenon of visual discomfort when viewing stereoscopic images,which reduces the viewing experience.Therefore,it is of great significance for cognitive research on the effects of convergence adjustment on the visual discomfort of stereo images.In this paper,we used the EEG technology to study the visual comfort caused by single parallax,and the characteristics of two types of EEG signals caused by parallax were extracted.After the Hilbert-Huang transform was performed on the objective electroencephalography(EEG)signal,the corresponding spatial domain features,frequency domain features and time domain features were extracted.The features were sorted by CPSorting algorithm and Relief F algorithm respectively.We used the CPSorting algorithm to perform spatial domain feature screening,and the Relief F algorithm to perform frequency domain feature and time domain feature screening,Then we used the Adaboost classifier to classify.The classification results showed that the best evaluation index is ?/? of the P3 electrode in the top zone.Through the observation of EEG topographic map,we found that with the increase of parallax,?/?in the uncomfortable state shows an upward trend in the forehead area of the temporal region and its periphery.At the same time,we found that the sample entropy of the cross-parent non-comfort state is the smallest,and the EEG signal is more unstable,which is more likely to cause the visual discomfort result in the subjective assessment.On this basis,this paper further studied the visual comfort caused by the change of parallax.We collected the Event Related Potentials(ERPs)and EEG signals of different parallax change levels,and compared the N200,P3 a and P3 b components in the ERPs signal under different conditions,and found that with the increase of task level,the amplitude of N200 and P3 b component increase.Through the analysis of Event-Related Spectral Perturbation(ERSP),we found that the uncomfortable state induces stronger low frequency activity.By constructing the brain function network through the characteristics of EEG signals,we found that as the level of parallax changes,the brain network becomes more complex,the degree of information exchange between the brain regions is obviously enhanced,and the more active regions are mainly distributed in the forehead area,the top area and the reclamation area.At the same time,as the frequency band increases,the differences in brain function network characteristics of the three parallax variation amplitudes gradually decrease.Finally,we extracted the corresponding network features,and found that the brain networks of the three states have good separability through the SVM classifier,and the average classification rate is 83.33%.