Abstract: Objective To explore the effect of band power and wavelet packet entropy in the recognition of hand imagery.Methods The data gained from brain computer interface competition in 2003 provided by Graz University of Technology.The electroencephalogram(EEG)signals between 8～16 Hz were decomposed by db3 wavelet packet at three levels.The band power(BP)and wavelet packet entropy(WPE)of C₃ and C₄ were calculated respectively.The BP and WPE were defined as the feature vector.The left and right hand motor imaginary tasks were distinguished.Results The proposed method was applied to the test data set with 140 trails.The satisfactory results were obtained with the highest classification accuracy 87.14%.Conclusion The band power and wavelet packet entropy of EEG changed with time is coincident with event-related desynchronization and event-related synchronization.It can be used to recognize the left and right band motor imaginary tasks.

Key words: electroencephalogram signals, band power, wavelet packet entropy, feature extraction, classification