Real time extraction of visual evoked potentials

On-line Information maximization (Infomax) algorithm for blind signal separation is based on single observation sample, it can track real time change of mixing system adaptively. To overcome the problem of misadjustment in steady state, we present an improved on-line Infomax algorithm with preprocessing rule of 2nd and 4th order statistical decorrelations, adaptive step size technique, and we apply them to real time extraction of Visual Evoked Potentials (VEP). The results of simulation and application experiment show that the proposed method improves the convergence performance and effect of blind signal separation. This algorithm can perform real time reinforcing and parameter extraction of VEP effectively on the premise of smaller vision stimuli.