Robot motion control using Brain Computer Interface

In this work, two dimensional motions of a robot are controlled using brain computer interface. Motor Imagery signals for different mental activities are recorded using electroencephalography technique. Recorded electroencephalogram signals are filtered out for noise reduction and processed. Processed signals are further used to prepare the feature vector to train classifier algorithm. Appropriately trained and tested classifier algorithm is used to translate Electroencephalogram signals to a meaningful command. These commands are the electrical signals which further control the two dimensional motions of robot. For implementing Brain Computer Interface, Electroencephalogram signals are filtered out by Butterworth low pass filter and further preprocessed using Multi-scale Principal Component Analysis algorithm. Support Vector Machine classifier completes classification task.