A thermal image based sleep summarization system

• Main Authors: Wei-KaiHung, 洪煒凱
• Other Authors: Pau-Choo Chung
• Format: Others
• Language: zh-TW
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/757a52

碩士 === 國立成功大學 === 電腦與通信工程研究所 === 105 === With social progress and medical development, the dramatic increase in average life expectancy is growing and the proportion of the elderly in Taiwan is increasing year by year. How to use start-of-art technology to improve the elderly’s daily life has become an important issue. With an aging population growing, the demand for nursing centers of the elderly has increased and nursing centers have been set up year after year. In order to avoid accidents, some of nursing centers would install cameras for monitoring the elderly’s conditions. However, it is necessary to keeps a person watching the screens all the time, which is a hung human burden. furthermore, sleeping plays a vital role in good health and well-being throughout elderly’s life. In order to understand the sleep of an elderly, caregivers have to spend their time checking the room regularly. Meanwhile, the intermittent between checking also become as a safety dead zone period. Hence, this paper presents a scheduled-based hotspot detection to detect activities of an elderly in the night, including going to bed and getting out of bed and develop a thermal image based sleep summarization system to automatically record detected results. The detected results could help caregivers not only understand the sleep situation of the elderly but also reduce their burden. Moreover, doctors could consider the information of our system when they prescribe medicines for the elderly. The experimental show that the proposed algorithms have satisfied detecting rate for the defined activities.