A Real Time Human Emotion Recognition System Using Respiration Inhale-Exhale Temperature Patterns and ECG

Abstract

A major requirement for human health monitoring in real time is the capability to evaluate their perceptual state. This paper presents a real time monitoring system for human emotions recognition using respiration inhale and exhale temperature patterns and ECG data, which is obtained by the sensors in real time interfaced with arduino. Our system is capable of evaluating three basic emotions (Happy, Sad, Fear) for adults and aged, using low cost and non-invasive physiological sensors. The sensors data is used by Neural Networks for training and classification of human emotions. To achieve automatic evaluation of the affective states of the human and to detect emotions; a novel system which integrates state of art hardware and software system is developed. The anatomical variables considered here actually have a very strong association with the changes in state of emotions with the subjects in our experiments as indicated by results. The experimental results shows the accuracy achieved approximately 75%, 71.87 and 70% for emotions happy, sad and fear in adults and 62.5%, 70% and 75% in aged people.