The Study Of Brain Computer Interface Based On Chinese Language Imagination

With the development of the field of Human-computer interaction,the applications of Brain-Computer Interface(BCI)gain its population in the field of games,medical treatment and military.Currently BCI signals are mainly about the non-invasive brain electrical signal,which can be easy to get and applied conveniently.As the BCI studies go further,experiments of various forms have been performed,mainly about induced BCI and Spontaneous brain interface.This article is to explore a simple experimental form of easy-to-operate model of the computer interface based on the spontaneous BCI.In this paper,a spontaneous BCI model based on Chinese language imagination is designed.We mainly study on the design and verification of experimental paradigm,the feature extraction and classification of offline data and the design of online experiment.The main contents of this dissertation are as follows.(1)By the study of the Chinese language processing model of the brain,functional brain regions are selected and experimental paradigms based on the Chinese language imagination are designed.Six Chinese phonetic initials and four Chinese characters were selected for language imagination experiment.The feasibility of the experimental paradigm was validated by analyzing ERG / ERS,single-channel data and brain functional regions of EEG signals.(2)In the data processing of offline experiment,principal component analysis(PCA)is selected to reduce the dimension of the data,and the relative energy of wavelet packet basis of multiple frequency segments is selected by wavelet packet transform as the feature of linear discriminant classifier.The results obtained are as follows: Pinyin language imagery and rest signal classification accuracy rate is77.24%,the classification accuracy rate of two pinyin imagery signals is 69.85%,the classification accuracy of three pinyin imagery is 50.76% The rate is 74.92%.The classification result of the two Chinese imagination signals is 68.69%.The classification accuracy of the three Chinese characters is 52.12%.(3)In this paper,an online experiment platform is built,and the corresponding channel is selected for online experiment through the index evaluation function of the established channel.The results of two online experiments of the same subject were56.63% and 61.96% respectively.In this paper,the research scheme of Chinese language imaginary brain machine interface is determined,which lays the foundation for the application of Chinese language imagination brain computer interface.