Using Fuzzy Adaptive Threshold Mechanism for Intelligent Assisted Detection System in Indoor Environment

碩士 === 南臺科技大學 === 資訊工程系 === 105 === With the increasing number of senior citizens, we have to pay attention to the healthcare system. Accidental falls are one of the leading causes of death in senior if proper measures are not taken in time after the fall. Consequences of inadequate care after a ser...

Full description

Bibliographic Details
Main Authors: CHEN, KAI-HONG, 陳凱鴻
Other Authors: HORNG, GWO-JIUN
Format: Others
Language:zh-TW
Published: 2017
Online Access:http://ndltd.ncl.edu.tw/handle/sgc6xg
Description
Summary:碩士 === 南臺科技大學 === 資訊工程系 === 105 === With the increasing number of senior citizens, we have to pay attention to the healthcare system. Accidental falls are one of the leading causes of death in senior if proper measures are not taken in time after the fall. Consequences of inadequate care after a serious accidental fall could include the senior having psychological trauma and thus lead to reduced physical activity. The healthcare system have a smart watch that can monitor the physiological values and fall detection, the Beacon device of indoor positioning, and sensor-based outdoor air quality equipment. A cloud server will be set as the system center in the Amazon, an APP will be designed to track position, physical condition, medication status, and reminder in the right time in the Android platform. In part of fall detection on smart watch, this study proposes an adaptive threshold algorithm based on fuzzy theory to proceed the fall determination. When the fall occurs, Beacon will immediately locate the position of the patient and send information through WI-FI to the server, and a notification will be sent to the caregiver phone apps through the server. During the experiment, we collected the acceleration data from 12 subjects (6 males and 6 females) in daily activities and accidentally falls to compare the differences between different height and weight in the acceleration values. Finally, we achieve 93.75% sensitivity and a specificity of 97.5%. Through this system, the existence of healthcare institutions can get more real-time information and feedback, and thereby enhance the overall quality of service and efficiency, and reduce labor costs and resource consumption.