Preliminary Study Of The Acquisition And Feature Extraction Methods Of Left And Right Hand Movement Imagery EEG

The advancement of Brain-Computer interface (BCI) study brings hopes of rehabilitation to the handicapped. And the BCI based on electroencephalograph has been a hotspot for the BCI study, because of its simple structure, safety and convenience of signal acquisition and portage.In this thesis, the characters of the Electroencephalogram (EEG) signal, the concept of the BCI , the domestic and international research status of BCI were introduced first. Then the paradigm of our own was groped for after the paradigm of the movement imagination studies of the groups oversea being studied, and the program'Visual Guide'for visual cues of the hands movement imagination was made. Testees were trained with this program and the signal with enough distinction was acquired finally with the successful rate of 83%.For the goal that features of the signal would be extracted effectively, the signal acquired previously was pretreated, and then the signal was analyzed in frequency domain , time domain and time-frequency domain. Through frequency domain analysis, the signal got from C3 and C4 had greatest distinction about the 10Hz, and the signal of 10Hz had correlation with the cues of left hand and right hand movement imagination through time domain analysis. And the time-frequency analysis had further sustained the conclusion. With the features that we got from the time-frequency domain analysis, the original signal was decomposed and the signal with the features was extracted successfully through the Morlet wavelet transformation.