The invention discloses a visual brain-computer interface method based on field-programmable
gate array local
noise optimization. The method comprises the following steps: placing an acquisition
electrode in a head pillow area to acquire a visual electroencephalogram
signal; generating a chessboard stimulation unit with local
noise, and storing the chessboard stimulation unit in an internal memoryof the
field programmable gate array; displaying the chessboard stimulation unit with the local
noise through a
field programmable gate array control mode; enabling a user to selectively
gaze at anychessboard stimulation unit with local noise, collecting electroencephalogram signals of a visual pillow area of the user to carry out
canonical correlation analysis, obtaining the visual stimulationunit corresponding to the maximum value in
canonical correlation coefficients corresponding to all frequencies, and judging the visual stimulation unit as an identified target. According to the invention, the display stability of the chessboard visual stimulation unit can be improved, and the requirements of the steady-state visual
evoked potential brain-computer interface
system on hardware are reduced, so the implementation cost of the brain-computer interface
system is reduced, and the recognition accuracy and stability of the steady-state visual
evoked potential brain-computer interface can be improved.