Brain computer interface (BCI) has become a new communication mode from brain to output device independent of normal peripheral nerves and muscles. EEG signal processing is one of the key technologies. In this paper, wavelet transform and empirical mode decomposition are used to decompose EEG signals to reconstruct eight IMF. Then the permutation entropy is calculated for the decomposed IMF. The IMF entropy with small feature redundancy is selected by using P value, and the support vector machine optimized by genetic algorithm is used for classification. The accuracy rate reaches 97.64%.