ANN Based Feature Extraction from EEG test Signals in Generating Control Commands for Prosthetic Application

The field of prosthetics has been emerged with technological advancements especially incorporate with signal processing controls. An enhancing feature in the field of prosthetics is the involvement of brain computer interface (BCI). This paper is used to generate control commands in prosthetics using signals. The work focused over the feature extraction from eye blinks from test data accompanied by neural networks. Ocular pulses pointed towards the signal waves of motor imagery actions were calculated with respective of their height and width. A user defined input and target data was developed from test data. Classification of the blink features was performed in machine learning and deep learning techniques. The measured sensitivity of the user data classification is 89%on KNN, 89.6%on SVM and 98.1% in ANN. The classification results by ANN shows the best effectiveness in generating controls for prosthetic applications