There are several methods of psychophysiological data collection from humans such as, Electrocardiogram (ECG), Galvanic Skin Response (GSR), Electromyography (EMG), and Electroencephalography (EEG). This paper is presenting the emotion recognition of EEG brain signals using Support Vector Machine (SVM). The emotions were elicited in the subjects using emotion related stimuli. We used the emotional stimuli from the International Affective Picture System (IAPS) database in this research. These

more »

... muli belonged to five types of emotions in our experiment such as, happy, calm, neutral, sad and scared. The raw EEG brain signals were preprocessed to remove the artifacts. We introduced a feature extraction method using Hjorth parameters. The set of features were extracted from preprocessed EEG signals of each subject, separately. The combined feature set of all subjects was processed through SVM. The results had shown the 70 % accuracy of emotion recognition in arousal-valence domain over 30 subjects. The classification result of all emotions was about 30% but we can see the improvement in accuracy over reduction of size in emotional group and classification accuracy of 70% is getting better up to two emotions.