Emotion analysis through physiological measurements.

There have been several advances in the field of affective computing analysis, however one of the main objectives of the brain computer interfaces (BCI), it is to interact in a natural way between human and machines; The analysis of the emotional state user is very important since it may provide a more suitable interaction generating closest approach, and efficient interaction. Furthermore extent that implementation of systems that can develope an emotional human-machine interface, applications of affective computing, it could widely used to help people with physical or deliberately to analyze and characterize the emotions that may be of interest and provide benefits on a human activity (eg, performance of athletes). In this paper seven emotional related experiments where development according whit the model arousal / valence model, each experiment corresponds to an evoked emotion from 32 study subjects and contains information the galvanic skin response (GSR), an electrooculogram (EOG) and electromyogram (EMG), for each user, each emotion is evoked by a process audio / visual for 60 seconds, focused generate three emotions (anger, happiness, sadness). Statistical measures are used to create parameters and distances to generate an overview of classification of signals related to an emotion. For information preprocessing is used discrete wavelet transform and statistical parameters (mean, standard deviation, variance) plus a surface filter.