Wavelet Based Classification of Finger Movements Using EEG Signals

Brain-computer interfaces (BCIs) have been examined in the field of bio-medical engineering. This braincomputer interface method is very useful for the people who are suffered by some nervous disorder to control or operate the external devices. EEG dataset are acquired and these signals are processed for identifying the brain thoughts to control the device. Here we proposed the method for the classification of the finger movements using EEG signals which are used in the application of artificial upper limb. This method includes pre processing, feature extraction and feature classification. Pre processing includes the removal of artifacts in the EEG signals due to some noises like eye blinking, etc. Discrete Wavelet Transform is used for the feature extraction. Features in both time domain and frequency domain are evaluated on the various EEG Signals. Based on the features extracted from various EEG signals, they are classified for the different finger movements using SVM classifier. The accuracy of 96.67% has been achieved for the proposed finger movement’s classification.