EEG features extraction and classification of rifle shooters in the aiming period

A basic problem in the design of EEG signal based devices, which could help the upper limb disabled soldiers carrying on their shooting tasks, is presented by the extraction and classification of EEG features. Such system can extract EEG signals features during soldiers act their shooting tasks and transform the features into binary control signals for operation. This paper is about analyzing the EEG signals of health soldiers during their rifle practice during the aiming period, which is the most vital step for shooting and extracting EEG features. We put the special features into a support vector machine to classify two classes signals and compare the signals of the holding period with an aiming period. Results show that the power of alpha and beta in occipital and parietal regions have significant changed, so does the power of theta rhythm in frontal area. Thus, we put the combine of alpha and beta power which as EEG features into our support vector machine’s classification device, then get the accurate classification rates compare with the one that comes from theta power. The alpha and beta power join as the characters get higher classification accuracy than the theta.