The invention relates to a
hand motion automatic correction and recognition method based on electroencephalogram
signal detection. The method includes the following steps that 1, ERP components of electroencephalogram signals of a healthy subject are acquired; 2, electroencephalogram signals, in a corresponding brain area, of a patient suffering from serious
motion limitation are acquired; 3, a feature calculation time window Fw is set, the
peak value, the mean value Mean, the standard deviation SD, the
correlation coefficient CORR of collected
electrode signals, the autoregression model coefficient AR and the electroencephalogram
signal energy E, in the feature calculation time window Fw, of the electroencephalogram signals in the finger motion automatic correction test of the healthy subject are acquired, and the
peak value, the mean value Mean, the standard deviation SD, the
correlation coefficient CORR of collected
electrode signals, the autoregression model coefficient AR and the electroencephalogram
signal energy E, in the feature calculation time window Fw, of the electroencephalogram signals in the finger motion automatic correction test of the patient suffering from serious limitation are acquired; 4, classification and recognition are carried out through a
binary classification support vector machine, and whether a motion completed by the subject in the
hand motion automatic correction test is a correctly-completed automatically-corrected motion is judged. Compared with the prior art, the method has the advantages of being applicable to
hand motion automatic correction, accurate in recognition result, broad in application prospect and the like.