Classification of motion direction perceived based on EEG

In this paper, an experiment was designed to record the electroencephalography when people caught the vision of differently directional moving (turning right, turning left, moving forward, moving backward). The signals from 30 EEG channels were analyzed by independent component analysis to obtain the independent components. The tagged permutations of independent components were obtained and a discrete space was formed by the tags. A discrete particle swarm optimization algorithm was used to search the optimal permutation through the discrete space. Based on the optimal permutation of independent components, the features were obtained and a high classification accuracy rate can be reached.