Falling detection of lonely elderly people based on NAO humanoid robot

Abstract

In recent years, elderly people are suffering an increasingly severe problem in their home due to the falls. Falling down is the most common cause of death among the elderly. This paper proposed a falling detection algorithm based on a humanoid robot to monitor and detect the motion of the elderly who live alone. A stream of the videos obtained from the robot cameras are transmitted to the personal computer via Wi-Fi, and then the image data will be analyzed by the MATLAB procedure algorithm, including extracting foreground, eliminating noise from foreground, evaluating Motion History and determining ellipse statistical properties to determine whether a falling has happened. If a fall occurs, the robot will raise an alarm to call for help through microphone. This algorithm will help the elderly receive timely medical treatment in order to minimize the falling risk and the company of NAO will enhance the elderly's sense of security.