Multi-Sensory Assistive Living System for Elderly in-Home Staying

One of the Ambient Intelligence (AmI) systems nowadays, is the Assisted Living System (ALS) that helps elderly people to live in an intelligent and pro-active environment. It is charged with the supervision and control of the entrusted space, monitoring events and human well-being through a network of heterogeneous sensors. This paper shows the capabilities of depth images extracted from RGB-D Microsoft Kinect camera fixed on the wall and 3-axis accelerometer concealed under smart tiles to localize and track elders, recognize their main Activities of Daily Living (ADLs) (walking, standing, sitting, lying down and the transitions between them) as well as detect falling down cases. While the proposed approach remains simple and non-intrusive, the experimental results show that it is able to achieve the aim with a high precision (sensitivity more than 95%).