Research On Single Channel Brain-computer Interface Technology Based On Artifact Removal

Brain-computer interface technology can transmit information between brain and computer or other devices.This technology can infer people's different thinking and cognition state by extracting and recognizing EEG signal,and then convert it into instruction to control the external devices.Brain-computer interface technology has broad application prospects in fields such as medicine,traffic,military,and smart home.According to the way of collecting EEG signal,brain-computer interface can be classified as invasive and non-invasive.And according to the number of electrodes for collecting EEG signal,non-invasive brain-computer interface can be divided into multi-channel and single-channel.The multi-channel acquisition mode is based on multiple acquisition electrodes to obtain low-distortion,high-quality EEG signal.However,it is necessary to apply the conductive gel to reduce the contact resistance of scalp when it is used.And it is generally suitable for areas where need highly quality of EEG signal.Single channel acquisition mode needs only one acquisition electrode,although the signal-to-noise ratio of EEG signal collected is very low.But it does not need to be coated with conductive gel,and it is easy to use and suitable for portable device development.Aiming at the demand of portable devices,a single channel brain-computer interface system based on artifact removal is designed.The system mainly includes EEG acquisition,EEG preprocessing and EEG transmission unit.In the process of system design,the wavelet transform is used to extract EEG feature signal,and achieve “mind control” of external devices.At the same time,through the artifact removal algorithm based on independent component analysis,the electrooculogram artifacts which overlap with the low frequency part of the EEG signal are effectively removed,and the stability of the system is improved.In order to verify the feasibility and effectiveness of the system and artifact removal algorithm,the test platform of car is built in this system.The actual test shows that the average success rate of ten volunteers' controlling of car is 80%.The system designed in this paper can satisfy the system performance requirements.In addition,the control success rate after artifact removal is 9.3% higher than that without artifact removal,which verifies the effectiveness of the artifact removal algorithm.