Privacy-Preserving Automatic Slipping Detection Method for Elderly in Bathroom Using Depth Sensors

Slipping is one of the main factors affecting the health of the elderly, especially in the humid bathroom environment. Failure to detect the elderly slipping in time may lead to more serious consequences. However, due to the privacy of the bathroom, it is not feasible to use ordinary video surveillance methods to monitor the elderly slipping in real time. In this research, we propose a slipping detection method based on depth sensors, which gets rid of the dependence on the video surveillance, can automatically detect the elderly slips and issue an alarm. First, based on the data collected by the depth sensors, we constructed a 3D scene point cloud representation without privacy information. Then, a pre-trained three-dimensional human body detection model is used to detect the accurate position of the human body. Finally, according to the coordinates of the 3D bounding box of the human body, it is judged whether the elderly has slipped. After verification on actual data sets, we achieved an accuracy rate of 98.4%, which not only ensures the privacy of the elderly, but also realizes timely slip alarm.