원문정보
초록
영어
Physiological information can objectively reflect the emotional state. Emotion recognition based on physiological signals has the important practical application value. In this paper, six kinds of chaotic characteristics were extracted from physiological data about Electromyogram, Skin conductivity, Respiration, which come from Augsburg university in Germany. Six kinds of chaotic characteristics contain the largest Lyapunov exponent, correlation dimension, box dimensions, information entropy, approximate entropy and complexity. On the basis of this, the degree of relation between chaotic characteristics of physiological signals and emotions was researched. The results showed that different chaotic characteristic values of different physiological signals sensitive to the emotion is different. Therefore, in order to improve the recognition rate, the chaotic characteristics of the multiple physiological information need to be merged to establish chaos matrix, which provide a new method for emotion recognition.
목차
1. Introduction
2. Selection of Physiological Signals
2.1. Electrocardiogram
2.2. Electromyogram
3. The Source of the Physiological Data
4. Extraction of Chaotic Characteristics Parameters
4.1. Extraction of the Largest Lyapunov Exponent
4.2. The Extraction of Correlation Dimension
4.3. The Extraction of Chaos Box Dimension
4.4. Extraction of Information Entropy
4.5. The Extraction of Approximate Entropy
4.6. Extraction of the Complexity
5. Establishment of Chaos Characteristic Matrix
6. Conclusion
References