The Development of Emotion Recognition System Using Home-based Physiological Signal Measurement

• Main Authors: Tzu-Yu Huang, 黃子瑜
• Other Authors: Yen-Ting Chen
• Format: Others
• Language: zh-TW
• Published: 101
• Online Access: http://ndltd.ncl.edu.tw/handle/49452376625190099564

碩士 === 南台科技大學 === 電機工程系 === 100 === Abstract As the people living in the modern society, there are many extrinsic factors could result the feeling of pressure. The factors might cause from family, personal economy, work, et, al. The negative emotion, such as depression, nervousness, anger, sadness, would be easily induced as people can’t relieve the pressure. Based on the theory of emotion, there is high causality between negative emotions and physiological events of human beings. Therefore, how to understand and relieve the personal emotion is an important topic for modern people. In this study, a home use chip-based physiological signals measurement device has been developed for measuring the physiological signals including electromyogram, electrocardiogram, body temperature, galvanic skin response, and photoplethysmogram. The measured physiological signals are measured and digitized, and then save to SD card. The LCD board is used for information displaying. The pictures released from International Affective Picture System(IAPS) were applied in this study to evoke the emotion of subjects. The feature parameters were then evaluated from the measured physiological signals. The results of statistical discriminant analysis showed that the difference is significant between the four kinds of emotions including happiness, sadness, fear, and disgust for the normal people and the people with depression-tendency (The score of HAM-D >7). And the difference is significant as well between the two groups. The device and software program can be set up for daily activities monitoring of patients. It can be implemented for patients with psychiatry disease to facilitate the tracking for curative effects for long-term treatment at home.