Research On Scene Reconstruction Technology Based On EEG Analysis

Being in a negative mood for a long time will have a lot of adverse effects on people’s life and work,and even cause harm to people’s health.Therefore,the regulation of negative emotions is particularly important.In the past,the most common way to relieve negative emotions was to use music and pictures.However,the traditional music and pictures lack the pertinence of individual to relieve the negative emotions,which leads to the unsatisfactory effect.Compared with music and pictures,the video has a certain plot and a strong sense of substitution.Therefore,this thesis puts forward the research of scene reconstruction technology based on EEG,in order to reconstruct a more soothing video.This study is divided into two parts: the identification of negative emotions and the alleviation of negative emotions.In the research of relieving negative emotions,the accurate recognition of emotions is very important.EEG is closely related to the change of emotion.In order to improve the accuracy of emotion recognition,this thesis constructs a HMM-SVM model based on Support Vector Machine(SVM)and Hidden Markov Model(HMM).The simulation results show that the average recognition accuracy of HMM-SVM is 84.88%,which is 3.56% higher than that of SVM.In view of the unsatisfactory effect of music and pictures,this thesis proposes to use reconstructed video to relieve negative emotions.In order to reflect the emotional information conveyed by music more accurately,this thesis uses sentence level features as music features.On the other hand,in order to achieve the high matching of music and image emotional information,this thesis proposes a scene reconstruction model.Through LPP algorithm,music features and image features are transformed and mapped to the same low dimensional subspace for comparison and effective fusion,which matches the correct emotional pictures for different music and achieves the goal of scene reconstruction.In order to further prove the soothing effect of scene reconstruction video,Self-assessment Manikins(SAM)and PSD change of EEG signal were used to evaluate the soothing effect.SAM score results show that the average intensity of pleasure emotion immediately after watching the scene reconstruction video is increased from 6.4 to 7.2,the emotional activity is increased by 2.3 points compared with that before the experiment,in addition,the scores of dominance and liking were improved;The feedback of EEG PSD after watching the reconstructed video showed that compared with the PSD in sad state,the maximum value of PSD in β-band was relatively weakened,and the PSD value inα-band was significantly strengthened,which increased from 350 μHWz/ to2200 μHWz/,and the brain activity became higher.Therefore,the scene reconstruction video based on EEG analysis has a positive effect on alleviating negative emotions.