Research On SSVEP Based BCI And Its Application In Telephone Communication And Character Spelling

Brain-computer interface (BCI) is a new communication system base on the machine-computer interface (MCI) that translates electrical brain signals into machine control command, without using traditional pathway consisting of peripheral nerves and muscle. Nowadays, BCI has entered into a practical research stage from the theoretical research stage. We seem the BCI practical as the target, this paper has builded a BCI Experimental Platform based on SSVEP. Online calling system and online spelling system also realize on the BCI experimental platform.The BCI Experimental Platform includes visual stimulator and online signal processing system. There are two kinds of visual stimulator:LED visual stimulator and LCD visual stimulator. Online signal processing system consists of system initialization, EEG signal online extract and signal processing, serial port communication. Online system is a software platform based on VC++which extracts and recombine the EEG to get the target channel data. The system used the PSDA analysis methods to realize feature extraction and recognition and send control command via serial port.This paper introduces two kinds of SSVEP signal feature extraction method, power spectral density analysis (PSDA) and canonical correlation analysis (CCA). PSDA is based on discrete signals Fourier transform method. CCA method find the max correlation coefficient between EEG and reference signal to extract the frequency feature.Online calling system and online spelling system are two Practical applications on the SSVEP BCI experimental platform. The online experiment result shows that two system get considerable recognition accuracy rate and Information transformation rate (ITR),...