Electroencephalogram signal denoising method, device and system based on empirical mode decomposition and Kalman filtering

The invention discloses an electroencephalogram signal denoising method, device and system based on empirical mode decomposition and Kalman filtering. Aiming at signals collected by a head electrode, the self-adaptive capability of processing non-linear and non-stationary signals through empirical mode decomposition is exerted, self-adaptive decomposition is carried out on electroencephalogram signals to obtain intrinsic mode components, and the integrity of local information is guaranteed; performing classification processing on the plurality of intrinsic mode components according to a correlation criterion; a Kalman filtering algorithm is used for carrying out de-noising processing on components, with the second highest similarity with the original electroencephalogram signals, in the intrinsic mode component signals, the problem that Kalman filtering diverges under a nonlinear system is avoided, meanwhile, the calculation speed is increased, and finally the electroencephalogram signals with noise removed a