Hitoshi Matsuyama
Nagoya University, Japan
Title: Designing Elderly Monitoring System by Fall Detection Method Using Inertial Sensor of Walking Stick
Biography
Biography: Hitoshi Matsuyama
Abstract
The number of old parents left in rural areas is increasing due to the diversification of lifestyles and outflow of their children into the city. Therefore, we need techniques and tools for monitoring the living conditions of the elderly even if they are away from their parents to protect them from health risks, especially falls, one of the most serious risks. Until now, sensors built in smartphones were generally used for fall detection. On the other hand, although smartphones are widely used by the elderly, they often go out without their smartphones. Therefore, we focused on walking sticks because the elderly people at a level that needs to be monitored when going out have a high carrying rate of walking sticks, and started to work on developing a fall detection method using stick attached by sensors. As the preliminary research, we attach a wearable sensor to the stick and collect acceleration and angular velocity data of activities. The fall and other anomalous activities are detected by stick fall recognition. Therefore, we develop the stick fall recognition method using Random Forest Classifier with sliding window metrics and signal processing. Adopting the developed recognition system, we design the monitoring system over the elderly and other people using walking sticks. The system collects the sensor data, uploads them to the server and after the analysis of the activities, it provides alert to the recipient such as children via the smartphone app when anomalous activities of the people using sticks occur.