The world is experiencing an unprecedented, enduring, and pervasive aging process. More and more elderly will be staying alone at home, especially in the daytime, and need to handle daily living activities by themselves. It is important to have effective measures of their self-caring abilities and to minimize accident risks. For example, an efficient fall detection method for the elderly, tracking their daily activities, is desirable. In this chapter, we discuss how to apply the latest intelligent sensor technologies to track the common indoor activities performed by an elderly person in his or her living quarters, which could be used fall detection. Through the introduction of our system SmartMind, we first show how Kinect, a 3D depth camera, can be applied for effective activity tracking of the user within a predefined environment. In the design of SmartMind, in order to improve accuracy in activity detection, we adopt a context-based approach to model the activities. Since Kinect has a privacy concerns problem, in the second part of the chapter, we introduce another system, called ActiveLife, in which simple motion sensors are adopted to measure changes in motion for indoor activity estimation. To improve its accuracy in activity estimation, we need a good model the living environment of the user and his/her activities within the living environment. Experimental results have shown the effectiveness of using a machine learning method, support vector machines, for improving its accuracy in activity estimation.
|Title of host publication||Intelligent Data Sensing and Processing for Health and Well-being Applications|
|Number of pages||26|
|Publication status||Published - 1 Jan 2018|
Scopus Subject Areas
- Activity tracking
- Machine learning methods
- Motion sensors