独居高齢者見守りのための深度カメラを用いた転倒検知システムの開発

―倒立振子モデルを用いた転倒動作学習用データの作成―

抄録

In Japan, the proportion of elderly people living alone is increasing, and there is a concern that in emergencies such as accidents, they may be unable to contact external help. Therefore, this study aims to develop a fall detection system for monitoring elderly individuals living alone. Using a depth camera to measure body acceleration, sudden accelerations during falls were confirmed. Furthermore, we created an inverted pendulum model in a dynamics simulator, conducted simulations of acceleration characteristics, and compared them with the measured acceleration during falls. As a result, it is considered that the inverted pendulum model is useful as training data for fall detection.