A Monitoring System to Support Home Health Care for the Elderly with Dementia by Detecting Going Out Activities Based on RGB-D Sensors

This paper proposes a monitoring system which can be used with health care robots. As an aging society is arriving, the interest of health care for dementia patients has been increasing. However, the problem of an aging society is that the number of care workers has been decreasing. In order to supplement this shortage, developing health care robots is a hot agenda including home health care service these days. In developing a dementia health care robot, it should be able to manage some difficulties of patients and emergency situations. Especially, the missing problem of people with developing dementia is serious because it can be directly connected to death. Thus, in this paper, we propose a monitoring system to prevent the disappearance of dementia patients in advance and, which can also support health care robots. For the implementation of this, three processing steps are included. First, a calibration method for multiple cameras is presented. In the second step, a simple image retrieval pipeline will be introduced for identifying people in a house. In the last, we will explain how to detect the situation of going out activities.