Abstract: Brain Computer Interface (BCI) is a communication pathway between a human brain and an external device. For paralysed people BCI acts as an interface to control and regulate external devices and replaces their lost motor functionality. A motor imagery BCI converts a subject’s thought about a motor activity into control signals which in turn controls the intended device. The EEG signals that are produced according to the motor imaging need to be processed and analysed using various signal processing algorithms. Learning and modelling the brain activity presents a huge challenge in the accurate classification of this EEG and hence affects the performance of the BCI system. The importance of feature extraction stage is that, since the brain regions work in collaboration during an activity the correlation between the EEG signals must be considered during this phase. The tasks need to be classified accurately by efficient feature translation algorithms. This paper is a study on different signal processing techniques for the accurate extraction of EEG features and their classification for an efficient motor imagery BCI system. Different discrimination algorithms based on frequency, temporal and spatial domains are being analysed.

Keywords: Brain Computer Interface, Linear Discriminant Analysis, Separable Common Spatio –Spectral Patterns, Neural Networks, Common Bayesian Network.